2026-10-15 · 週四 10 篇
- 1 👍 321
Any Time, Any Space: A World-State-Centric Framework for Agentic Video Generation
ATAS improves long-form video consistency by compiling screenplays into referenced shot requests, retrieving spatial and temporal evidence from scene and event …
- 2 👍 233
Beyond Domain-Routed Supervision: Cross-Teacher Guarding for Multi-Teacher On-Policy Distillation
Road MOPD reduces gradient conflicts in multi-teacher distillation by using cross-teacher agreement for token-level credit assignment and auxiliary gradients to…
- 3 👍 189
Player-1: Learning to Play from Synthetic Experience
Player-1 improves game control by combining supervised fine-tuning on gameplay trajectories with reinforcement learning across synthetic environments, achieving…
- 4 👍 92
Doc2Harness: Benchmarking LLMs on Long-Horizon Harness Engineering of Coding Agents
Doc2Harness is a deterministic two-stage benchmark that evaluates long-horizon harness engineering by having models build and run harnesses across isolated cont…
- 5 👍 72
More Methods, Better Agents? A Reality Check on Self-Distillation in Agentic Reinforcement Learning
Researchers benchmarked ten self-distillation methods against GRPO across multiple agent tasks and model sizes, finding limited consistent gains, hidden trainin…
- 6 👍 70
Recovering Reliable Supervision via Critique-Guided On-Policy Distillation
CriticOPD improves on-policy distillation by using teacher critiques to restore recoverability, enabling students to revise incorrect reasoning and achieve stro…
- 7 👍 64
LatentLoop: Scaling Looped MoE Models with Asymmetric Compute Allocation
LatentLoop improves looped Mixture-of-Experts by compressing the first loop visit, reallocating saved computation to larger hidden dimensions, and using separat…
- 8 👍 62
Cortex: Consolidating Context into Parametric Memory
Cortex consolidates context-induced improvements into mergeable low-rank or structural updates, preserving non-target capabilities across sequential edits and d…
- 9 👍 52
ActionOPD: Distilling Decisions, Not Tokens
ActionOPD aligns smaller language-model agents with teachers via semantic action-level on-policy distillation, improving out-of-distribution tool-use performanc…
- 10 👍 42
Seeing, Maintaining, and Learning: The Evolution of Multimodal Agent Memory
This work proposes a two-axis framework for multimodal agent memory covering information formation and adaptive management to support lifelong learning.
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Federated Generation and Fusion for Privacy-Preserving Cross-Domain Optic Disc and Cup Segmentation in IoT-Enabled Healthcare
Ocular diseases are a leading cause of permanent blindness worldwide, making early and accessible screening critical for vision preservation. Fundus imaging, in…
- 2 Semantic Scholar
CEGI-S: A Counterfactual Evidence-Gated Intelligence Framework for Reliable and Self-Correcting Large Language Models
CEGI-S: A Counterfactual Evidence-Gated Intelligence Framework for Reliable and Self-Correcting Large Language Models. International journal of innovative resea…
- 3 Semantic Scholar
Generative AI And Large Language Models in Intelligent Systems
Generative AI And Large Language Models in Intelligent Systems. International journal of innovative research in technology…
- 4 Semantic Scholar
Closed-loop semantic–visual adaptation for identity-preserving personalized image generation
Closed-loop semantic–visual adaptation for identity-preserving personalized image generation. Multimedia Systems…
- 5 Semantic Scholar
An evaluation guided multiagent framework for generating floor plans that comply with municipal building regulations
An evaluation guided multiagent framework for generating floor plans that comply with municipal building regulations. Discover Artificial Intelligence…
- 6 Semantic Scholar
A study on three-stage integration of generative artificial intelligence in architectural design based on human-in-the-loop
A study on three-stage integration of generative artificial intelligence in architectural design based on human-in-the-loop. Discover Artificial Intelligence…
- 7 Semantic Scholar
Diffusion-Collaborated Text--Visual Multimodal Guidance for Image Manipulation Localization in IoT Supply Chain Systems
AI-powered supply chain systems, a critical component of IoT, leverage automated image verification for deliveries but face vulnerabilities such as image manipu…
- 8 Semantic Scholar
A GenAI-Powered Authentication Protocol for Secure Monitoring of Hajj/Umrah Pilgrims Using Cinematographic Drones
Maintaining safety in crowded events with more than 4.5 million Hajj and Umrah pilgrims becomes a very challenging issue in terms of crowd security. While occlu…
- 9 Semantic Scholar
Impact Of Prompt Design on The Accuracy and Reliability of Generative AI Responses
Impact Of Prompt Design on The Accuracy and Reliability of Generative AI Responses. International journal of innovative research in technology…
- 10 Semantic Scholar
Ensemble LLM Consensus Architectures for Deterministic Trade Signal Generation: An Empirical Study of Multi-Persona AI Panels in Retail FinTech Applications
This paper presents an empirical study of a multi-persona large language model (LLM) consensus architecture deployed in a live retail FinTech platform for equit…
- 11 Semantic Scholar
AI-driven multi-agent digital twin conceptual framework for real-time domestic water quality monitoring, purification and smart billing
AI-driven multi-agent digital twin conceptual framework for real-time domestic water quality monitoring, purification and smart billing. Discover Internet of Th…
- 12 Semantic Scholar
Deep Learning-Based Signal Detection With Denoising and Temporal Sequence Structuring for Molecular Communication in Vessel-Like Channels
Molecular communication via diffusion (MCvD) in vessel-like channels is strongly affected by laminar flow, diffusion-induced dispersion, and severe intersymbol …
- 13 Semantic Scholar
Compliance-Aware, Human-Gated Cloud Remediation: An LLM-Augmented, Specificity-Scored Framework for Multi-Jurisdictional Cloud Security Posture Management
Compliance-Aware, Human-Gated Cloud Remediation: An LLM-Augmented, Specificity-Scored Framework for Multi-Jurisdictional Cloud Security Posture Management. Inte…
- 14 Semantic Scholar
Fractional-order dynamical modelling of neurophysiological reaction–diffusion model of adiposity-induced cognitive decline using physics-informed neural networks
Fractional-order dynamical modelling of neurophysiological reaction–diffusion model of adiposity-induced cognitive decline using physics-informed neural network…
- 15 Semantic Scholar
Student use of generative AI and perceived institutional support across three linguistic contexts
Student use of generative AI and perceived institutional support across three linguistic contexts. Discover Education…
- 16 Semantic Scholar
Engaging HBCU Students in Active Learning for Blockchain, FinTech, and Generative AI: A Human-Centered Pedagogical Framework
The convergence of blockchain, financial technology, digital assets, and generative artificial intelligence creates important opportunities and challenges for u…
- 17 Semantic Scholar
Practical Mimicry and Modal Integration: Know-How, Large Language Models, and Human–AI Collaboration
Practical Mimicry and Modal Integration: Know-How, Large Language Models, and Human–AI Collaboration. Philosophy & Technology…
- 18 Semantic Scholar
SEFRQO-Plus: a self-evolving query optimizer via large language models and retrieval-augmented generation
SEFRQO-Plus: a self-evolving query optimizer via large language models and retrieval-augmented generation. The VLDB journal…
- 19 Semantic Scholar
CASSIS: An LLM-Driven Framework for Sensitive API Identification in Cyber–Physical System Software Repositories
As cyber–physical systems (CPSs) are increasingly deployed in critical infrastructures, the misuse of CPS-sensitive (safety-critical) data can compromise device…
- 20 Semantic Scholar
Fed-DAPT: Federated Dynamic Alignment and Prompt Tuning for Efficient Personalized Vision Transformers
Federated learning (FL) enables multiple clients to collaboratively train a model without exchanging raw data, making it attractive for privacy-sensitive edge c…
- 21 Semantic Scholar
Mapping generative AI in higher education teaching and learning: a scoping review
Mapping generative AI in higher education teaching and learning: a scoping review. International Journal of Educational Technology in Higher Education…
- 22 Semantic Scholar
A Michigan style genetic fuzzy system framework for multiple sclerosis: towards a digital twin application
A Michigan style genetic fuzzy system framework for multiple sclerosis: towards a digital twin application. Applied intelligence (Boston)…
- 23 Semantic Scholar
A Unified System for Intelligent Financial Fraud Detection
Modern financial ecosystems require real-time, transparent, and trustworthy fraud detection systems capable of processing massive transaction volumes while meet…
- 24 Semantic Scholar
When synthetic Delphi panels work: a graph-native methodology for participatory foresight under structural complexity
When synthetic Delphi panels work: a graph-native methodology for participatory foresight under structural complexity. European Journal of Futures Research…
- 25 Semantic Scholar
Graph-augmented LLMs with risk-adjusted safety filtering for personalized advice and medication recommendation
Graph-augmented LLMs with risk-adjusted safety filtering for personalized advice and medication recommendation. Frontiers of Computer Science…
- 26 Semantic Scholar
ColorFuseNet: hierarchical priority-guided tri-stream diffusion for controllable garment image generation
ColorFuseNet: hierarchical priority-guided tri-stream diffusion for controllable garment image generation. Multimedia Systems…
- 27 Semantic Scholar
Development and Exploratory Evaluation of a Large Language Model-Supported Scoping Review Workflow
Development and Exploratory Evaluation of a Large Language Model-Supported Scoping Review Workflow. Journal of medical systems…
- 28 Semantic Scholar
Building Financial Resilience at HBCUs through Generative AI, FinTech, Blockchain, and Cybersecurity Education
Financial resilience has become an urgent educational priority as individuals face increasingly complex decisions involving budgeting, debt, credit, retirement …
- 29 Semantic Scholar
PaCR: An Agentic Preference-Aligned TAP Conflict Handling
With the increasing complexity of smart home environments, trigger-action-program (TAP) conflicts have emerged as critical issues, occurring when multiple autom…
- 30 Semantic Scholar
Automated optimization modeling and auditing via dual-team multi-LLM-agent cooperative framework
Automated optimization modeling and auditing via dual-team multi-LLM-agent cooperative framework. Applied intelligence (Boston)…
2026-10-14 · 週三 10 篇
- 1 👍 276
Jeva: System One Models from First Principles
System one models enable direct, low-latency decision-making. Yet recent practice of predicting an answer from serialized text does not, by itself, respect the …
- 2 👍 133
VideoJev: System One Model for Video Understanding and Interaction
Visual agents must turn video observations into timely decisions to track progress and choose what to do next. Building on the emerging System One paradigm, we …
- 3 👍 116
Urbex: Agentic Spatial Grounding for City-Scale 3D Scenes
City-scale 3D grounding is commonly formulated as candidate ranking, which requires precomputed instance proposals and becomes inefficient in large outdoor scen…
- 4 👍 100
Tev: Scaling Long-Horizon Agentic Environments with Dense Progress Rewards
Long-horizon agents have drawn growing attention for their ability to work autonomously -- from hours to days -- on economically valuable tasks, a capability es…
- 5 👍 89
Harness^2: Recursive Agent Harnessing for an Open World
LLM agents' capabilities depend not only on their underlying models but also on their execution harnesses. Recursive self-improvement (RSI) offers a promising a…
- 6 👍 88
GreenWAM: A World Foundation Model for Robotics
A robot needs a world model that predicts how its actions will change the scene and a policy that chooses those actions. Models are usually built and evaluated …
- 7 👍 78
TPPO: Tracking-Value PPO For Language Model Reasoning
Proximal Policy Optimization (PPO) is widely used in reinforcement learning with verifiable rewards for language-model reasoning. While value evolution along tr…
- 8 👍 75
WorkRover: Scaling Long-Horizon Cowork Agent Training Via Recursive Environment Evolution
Scaling interactive environments is important for training long-horizon agents through reinforcement learning. These environments often require further testing …
- 9 👍 71
Vela: User Behavior as Supervision Sparks LLM Intelligence
Large language models are increasingly deployed in applications that act on users' long-term behavior histories, from personal assistants to recommendation and …
- 10 👍 71
DoSA: Dots Sparse Attention
Sparse attention visits few keys, yet finding them can still require scoring every key with every indexer head. We introduce DoSA (Dots Sparse Attention), a los…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar 📖 1
AI literacy and STEM education: emerging frontiers and research priorities
AI literacy and STEM education: emerging frontiers and research priorities. International Journal of STEM Education…
- 2 Semantic Scholar 📖 1
End-to-End Contextualized URL Extraction from Scholarly Papers
Uniform resource locators (URLs) have become ubiquitous in scholarly papers, and extracting them is an upstream task for downstream analyses such as link rot an…
- 3 Semantic Scholar
Why Large Language Models will Never be Embodied: A Categorical, Not Engineering, Impossibility
Why Large Language Models will Never be Embodied: A Categorical, Not Engineering, Impossibility. Minds and Machines…
- 4 Semantic Scholar
Author Correction: Large language models are powerful electronic health record encoders
Author Correction: Large language models are powerful electronic health record encoders. npj Digital Medicine…
- 5 Semantic Scholar
Cross-lingual transfer for post-traumatic stress disorder detection from clinical interview transcripts via LLM-assisted Turkish translation and multilingual transformer fine-tuning
Cross-lingual transfer for post-traumatic stress disorder detection from clinical interview transcripts via LLM-assisted Turkish translation and multilingual tr…
- 6 Semantic Scholar
MIRAGE: manipulation-immune RAG with adaptive guard against textual evidence attacks
MIRAGE: manipulation-immune RAG with adaptive guard against textual evidence attacks. Cybersecurity…
- 7 Semantic Scholar
Hashtags of hope: technology, stakeholders and disaster response
This research examines how individuals exercised digital agency by reaching out to emergent actors on Twitter (now called as X), and how this reshapes understa…
- 8 Semantic Scholar
Fluent but not fail-safe: machine-translated radiology reports still need expert review
Fluent but not fail-safe: machine-translated radiology reports still need expert review. European Radiology Experimental…
- 9 Semantic Scholar
An explainable graph transformer framework integrating structural reasoning and visual attention for interpretable nail disease diagnosis
An explainable graph transformer framework integrating structural reasoning and visual attention for interpretable nail disease diagnosis. Discover Artificial I…
- 10 Semantic Scholar
ChatGPT’s feedback credibility and students’ perceived learning: the role of agent-related social perceptions
ChatGPT’s feedback credibility and students’ perceived learning: the role of agent-related social perceptions. International Journal of Educational Technology i…
- 11 Semantic Scholar
Skill-guided agentic reasoning enables longitudinal synthesis of myeloma records at expert-level concordance
Skill-guided agentic reasoning enables longitudinal synthesis of myeloma records at expert-level concordance. npj Precision Oncology…
- 12 arXiv
ChunkAct: Action Chunking for Agentic Robotics
Agentic robotics places a large language model (LLM) agent in the robot control loop, where it composes perception, motion, and vision-language-action (VLA) too…
- 13 arXiv
LoopWAM: Closed-Loop Generalist Policy Improvement from Deployment Rollouts
Scaling expert demonstrations alone does not enable generalist robot policies to improve from deployment. Real-world rollouts contain successes, suboptimal beha…
- 14 arXiv
Dog-Breed Text-to-image Flow Matching from Pretraining to Fine-Tuning on 4 Consumer GPUs
We generally perceive that training text-to-image diffusion models requires heavy compute: it is true if we aim at a comprehensive frontier model. This understa…
- 15 arXiv
What Matters in Quality-Aware VLA Pretraining: A Benchmark and Empirical Study
How should demonstration quality be used during vision-language-action (VLA) pretraining to improve performance after target-task fine-tuning? We compare weight…
- 16 arXiv
PLAYGROUND: Progressive Layout Generation with Render-Grounded Planning
Generating editable graphic designs requires arranging visual and textual elements into a coherent composition while preserving their content and editability. A…
- 17 arXiv
Data Free Training of Diffusion Models for Lattice Field Theories
Sampling equilibrium field configurations remains a major computational bottleneck in lattice field theory, motivating the development of generative samplers th…
- 18 arXiv
Hybrid Reasoning Agents
How can we build an agent that thinks beyond human language, yet still acts through it in the external environment? Modern LLM agents still reason and interact …
- 19 arXiv
Visual Tokens on Demand: Visual Budgeting for Scalable and Efficient Multi-Image Reasoning in Multimodal LLMs
Multimodal large language models have advanced rapidly, but most remain optimized for single-image inputs. Multi-image reasoning is costly because each image ad…
- 20 arXiv
PixelMoE: Scaling Pixel-Space Diffusion with LLM-Style Mixture-of-Experts
Pixel-space diffusion models provide an appealing alternative to latent-space diffusion models by directly modeling raw images without relying on a learned toke…
- 21 arXiv
Exact Topological Compliance: Generating Persistence-Equivalent Graphs
Persistent homology summarizes the topological evolution of a graph as persistence diagrams (PDs). Generating graphs whose topology matches a target descriptor …
- 22 arXiv
Two Clocks: Why Larger Rollout Batches Learn Slower but Fail on Schedule in Reinforcement Learning with Verifiable Rewards
Reinforcement learning with verifiable rewards (RLVR) relies on costly model-generated responses, motivating larger rollout batches and square-root learning-rat…
- 23 arXiv
MFG: Multi-turn Feedback Guided Molecular Optimization
Multi-objective molecular optimization is a critical process in drug discovery that aims to modify a given molecule to simultaneously improve multiple propertie…
- 24 arXiv
The Last Remote Sensing Model Built by Humans? Casting Any Earth Observation Task as Recursive Self-Improvement
Earth observation (EO) models are still built by hand. For each new task, experts design a training recipe and tune it against a benchmark. We show that this lo…
- 25 arXiv
RepProbe: Can Video Language Models Count Repeated Actions? A Diagnostic Benchmark
Repetition counting requires identifying and enumerating action cycles over time. While state-of-the-art video language models describe actions fluently, they f…
- 26 arXiv
LIBERO-RIG: A Hardware-Setup Robustness Benchmark for Vision-Language-Action Models
Robot policies are expected to generalize across deployment environments, including variations in the robot itself: cameras are bumped and remounted, calibratio…
- 27 arXiv
Who Improves the Robot? Toward Physical Recursive Self-Improvement
Robots are increasingly autonomous in task execution, yet their improvement remains largely human-driven. This study investigates Physical Recursive Self-Improv…
- 28 arXiv
Fast Rates for Offline Q-Learning with KL Regularization
This paper investigates offline Q-learning for finite-horizon Markov decision processes with Kullback-Leibler (KL) regularization. Recent studies have shown tha…
- 29 arXiv
Scaling Online Optimal Transport Minibatches for Flow Matching
A central challenge in scaling flow models is maintaining stability in training and integration. In flow matching, this stability depends critically on how sour…
- 30 arXiv
Plan-Bench: Can Existing VLMs Go Well Beyond Perception to Reason and Plan in Embodied Worlds?
Vision-Language Models (VLMs) have achieved strong performance in visual perception and spatial reasoning, leading to growing interest in using them as high-lev…
2026-10-13 · 週二 10 篇
- 1 👍 266
GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives
Precise grounding matters. It specifies which object is the target and where that object is, even in clutter and for tiny objects, and it has to be fast enough …
- 2 👍 222
Thinking Inertia: LLMs Keep Thinking When Told Not To
Large Language Models (LLMs) increasingly ship with explicit "thinking modes", yet their counterpart, "no-thinking", has received far less attention. We study L…
- 3 👍 206
AgentSWE: Can Coding Agents Build the Agent You Actually Want?
As the production of software passes into the hands of software itself, coding agents are increasingly commissioned to build agents fitted to their users' own w…
- 4 👍 140
Scaling Agentic Data for Long-Horizon Terminal Intelligence
Terminal agents are moving beyond local operations toward completing real-world professional tasks. Short trajectories can demonstrate commands and tool calls, …
- 5 👍 82
VIS-Ground: Video Interactive Storytelling with Contextual Grounding
Video interactive storytelling enables viewers to actively steer how a video unfolds. However, once we allow viewers to intervene during generation, a new chall…
- 6 👍 73
GroundAnything: Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed
Autoregressive (AR) grounding models serialize spatial predictions, introducing sequential latency and imposing a causal order on output tokens. We view groundi…
- 7 👍 35
Multi-Stream Attention Residuals
Attention Residuals (AttnRes) enables content-dependent retrieval over network depth, which can be viewed as single-stream softmax attention along the depth axi…
- 8 👍 30
PR-OPD: Privileged Representation On-policy Self-Distillation for Agentic Reinforcement Learning
Language-model agents are usually trained by reinforcement learning from one reward per episode, and privileged self-distillation enriches it by letting the sam…
- 9 👍 29
Relevance Does Not Imply Applicability: Experience Activation for Personal GUI Agents
Personal Graphical User Interface (GUI) agents rely on interaction history to infer what a user wants from ambiguous instructions and to anticipate recurring ro…
- 10 👍 25
CrossTimeEdit: A Decade-Spanning Cross-View Dataset and Reward-Guided Editing for Historical Street-View Generation
Historical street-view imagery records urban evolution, but uneven coverage leaves substantial gaps in historical records. Generating plausible past appearances…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Can LLMs Simulate Expert Panels? Evaluating CAT-Inspired Persona Prompting for Automated Creativity Scoring
Can LLMs Simulate Expert Panels? Evaluating CAT-Inspired Persona Prompting for Automated Creativity Scoring.…
- 2 Semantic Scholar
Does Retrieval Relevance Matter for AI-Generated Book Introductions? A Matched-Random Memory Evaluation
Does Retrieval Relevance Matter for AI-Generated Book Introductions? A Matched-Random Memory Evaluation.…
- 3 Semantic Scholar
From cyberspeak to algorithmic newspeak: soft derogation and the ethics of AI-generated language
This paper examines the ethical and epistemological implications of language generated by large language models (LLMs) through the lenses of speech act theory, …
- 4 Semantic Scholar
Towards clinical communication training with LLM-based conversational virtual patients and automatic evaluation systems
Towards clinical communication training with LLM-based conversational virtual patients and automatic evaluation systems. Discover Artificial Intelligence…
- 5 Semantic Scholar
Toward Contextual Mentorship Inference: Bibliometric Modeling and LLM-Informed Reasoning over Scholarly Publication Metadata
Toward Contextual Mentorship Inference: Bibliometric Modeling and LLM-Informed Reasoning over Scholarly Publication Metadata.…
- 6 Semantic Scholar
Structuring The Top-K Results: The Case for Narrative Extraction, Structuring, and Visualization in Scholarly Research Tools
Structuring The Top-K Results: The Case for Narrative Extraction, Structuring, and Visualization in Scholarly Research Tools.…
- 7 Semantic Scholar
Can LLMs Identify Advisor-Advisee Relationships in Coauthorship Network Based on Publication Metadata?
Can LLMs Identify Advisor-Advisee Relationships in Coauthorship Network Based on Publication Metadata?.…
- 8 Semantic Scholar
Auditing Declared AI Generative Crawler Traffic on a Large-Scale Multilingual Digital Library
Auditing Declared AI Generative Crawler Traffic on a Large-Scale Multilingual Digital Library.…
- 9 Semantic Scholar
An opinion evolution prediction model based on graph attention and ordinary differential equations
An opinion evolution prediction model based on graph attention and ordinary differential equations. Discover Artificial Intelligence…
- 10 Semantic Scholar
From Digital Library Records to Trust-Grounded Virtual Experts for Knowledge Retention
From Digital Library Records to Trust-Grounded Virtual Experts for Knowledge Retention.…
- 11 Semantic Scholar
Triaging Defects in Scholarly Result Graphs with Extraction, Evidence, and Consistency Signals
Triaging Defects in Scholarly Result Graphs with Extraction, Evidence, and Consistency Signals.…
- 12 Semantic Scholar
Structured Knowledge Graph for Scholarly Exploration: IR and QA System for ETDs and their Elements
Structured Knowledge Graph for Scholarly Exploration: IR and QA System for ETDs and their Elements.…
- 13 Semantic Scholar
Which Text Normalization Methods Matter for Example-Based Retrieval of OCR-Generated Japanese Classical Texts?
Which Text Normalization Methods Matter for Example-Based Retrieval of OCR-Generated Japanese Classical Texts?.…
- 14 Semantic Scholar
Supplemental generative AI tutoring improves financial modeling scores in a two class quasi experimental study
Supplemental generative AI tutoring improves financial modeling scores in a two class quasi experimental study. Discover Computing…
- 15 Semantic Scholar
Towards Automated Analysis of Scientific Limitations and FutureWork
Towards Automated Analysis of Scientific Limitations and FutureWork.…
- 16 Semantic Scholar
Same Passages, Different Readings: Comparing Complete Human and LLM Interpretations of a Story
Same Passages, Different Readings: Comparing Complete Human and LLM Interpretations of a Story.…
- 17 Semantic Scholar
ABSP: Attention-Based Semantic Preservation for Faithful Context Compression
ABSP: Attention-Based Semantic Preservation for Faithful Context Compression.…
- 18 Semantic Scholar
Toward Preserving Open Access Datasets and Software in Scholarly Documents
Toward Preserving Open Access Datasets and Software in Scholarly Documents.…
- 19 Semantic Scholar
The Decomposition Dilemma: When Do Multi-Agent LLM Workflows Help Grounded Knowledge Generation?
The Decomposition Dilemma: When Do Multi-Agent LLM Workflows Help Grounded Knowledge Generation?.…
- 20 Semantic Scholar
Channel learning and enhanced adaptive reconstruction for robust image transmission over complex time varying wireless channels
Channel learning and enhanced adaptive reconstruction for robust image transmission over complex time varying wireless channels. Discover Artificial Intelligenc…
- 21 Semantic Scholar
Understanding Researchers' GenAI Adoption: A Grounded Theory of Human-AI Collaborative Decision-Making
Understanding Researchers' GenAI Adoption: A Grounded Theory of Human-AI Collaborative Decision-Making.…
- 22 Semantic Scholar
Extracting and Structuring Experimental Knowledge from Electronic Theses and Dissertations
Extracting and Structuring Experimental Knowledge from Electronic Theses and Dissertations.…
- 23 Semantic Scholar
Stuck on the First Answer? Auditing Primacy Bias in Multi-Turn LLM Preference Revision via Causal Masking
Stuck on the First Answer? Auditing Primacy Bias in Multi-Turn LLM Preference Revision via Causal Masking.…
- 24 Semantic Scholar
Bridging the Physicality Gap: Conversational Repair Strategies in AI-Supported Maker Learning
Bridging the Physicality Gap: Conversational Repair Strategies in AI-Supported Maker Learning.…
- 25 Semantic Scholar
Quantifying the Digital Heritage: A UNESCO-Inspired Value Index for Software Archives
Quantifying the Digital Heritage: A UNESCO-Inspired Value Index for Software Archives.…
- 26 Semantic Scholar
ChemSafe: Hybrid Retrieval-Augmented Decision Support for Laboratory Chemical Hazard Assessment
ChemSafe: Hybrid Retrieval-Augmented Decision Support for Laboratory Chemical Hazard Assessment.…
- 27 Semantic Scholar
Contextual Document Discovery in Digital Libraries through Hybrid Summarization and LLM-Guided Query Decomposition
Contextual Document Discovery in Digital Libraries through Hybrid Summarization and LLM-Guided Query Decomposition.…
- 28 Semantic Scholar
Examining Dermatology Image Generation Capabilities of Generative AI Across Skin Tones
Examining Dermatology Image Generation Capabilities of Generative AI Across Skin Tones.…
- 29 Semantic Scholar
Adaptive LLM Selection for Question Answering over Historical Digital Libraries
Adaptive LLM Selection for Question Answering over Historical Digital Libraries.…
- 30 Semantic Scholar
Segmentation-then-Summarization for Long-Form Audiovisual Archives: A Comparative Study of Three Pipelines
Segmentation-then-Summarization for Long-Form Audiovisual Archives: A Comparative Study of Three Pipelines.…
2026-10-12 · 週一 10 篇
- 1 👍 271
S-Space: Exploring Spatial Workspace in Multimodal Language Models
Multimodal language models (MLLMs) exhibit promising spatial reasoning capabilities, yet the internal mechanisms supporting them remain poorly understood. We un…
- 2 👍 242
DiffusionART: Efficient, Stable and Scalable Reinforcement of Diffusion Models
Online reinforcement learning (RL) fine-tunes diffusion models with black-box rewards, but reverse-process methods such as Flow-GRPO are limited in efficiency b…
- 3 👍 138
Reflex: Visually Grounded Reactive Humanoid Control
Humans effortlessly react to the world; they adeptly return a tennis shot, catch a falling cup, step aside from a passing cart, or pass a box tossed during truc…
- 4 👍 92
SoftwareGym: Scaling Computer-Use Agent Training with Diverse Synthesized Software Environments
Computer-use agents (CUAs) are rapidly improving, yet synthesizing diverse and reliable software environments at scale remains challenging. Native application e…
- 5 👍 66
EmbodiedCLBench: Evaluating Continual Learning for Self-Evolving Embodied Agents
Continual learning requires models to accumulate past experience and transfer it to new tasks. Despite substantial research on continual learning, existing work…
- 6 👍 58
VisionWeave: Weaving Elastic Visual Representations as a Native Capability of MLLMs
Multimodal large language models have become the dominant paradigm for visual understanding, but incur substantial costs by encoding inputs into dense, fixed-si…
- 7 👍 54
Here the World in Stereo: Learning Dynamic Spatial Correspondence for Immersive Joint Video-Audio Generation
Recent joint video-audio generation models have achieved strong semantic correspondence and temporal synchronization. However, applications such as AR/VR and in…
- 8 👍 49
ProgramGym: Scaling Training Environments for Long-Horizon Program Reproduction Agents
Coding agents are increasingly used to build programs from scratch as software engineers do rather than to resolve issues within a repository or to generate a r…
- 9 👍 43
Work2Skill: Capturing Human GUI Experience for Reusable Agent Skill Induction
Computer-use agents can perform increasingly complex GUI tasks, yet unfamiliar software and workflows requiring human domain knowledge remain challenging. Human…
- 10 👍 42
Continually Post-training Small Models via Synthetic Self-distillation
Continual post-training faces a dual challenge of acquisition and retention: models must learn new capabilities while preserving prior knowledge. While Self-Dis…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar 📖 1
AS2Biz: Leveraging Web Presence and AI to Improve AS Business Classification
Classifying Autonomous Systems (ASes) by the business sectors of their operating organizations, also referred to as AS business classification, is a relatively …
- 2 Semantic Scholar 📖 1
Can You See Me, GenAI? Characterizing Video-Based Conversational Interaction with Generative AI
Video-based conversational interaction between humans and generative artificial intelligence (GenAI) represents a rapidly emerging paradigm at the intersection …
- 3 Semantic Scholar
Towards automated formal verification of authorization policy in microservice systems
Towards automated formal verification of authorization policy in microservice systems. Cluster Computing…
- 4 Semantic Scholar
From prompt to profession: how ChatGPT usage relates to students' perceived employability through skill development
The proliferation of generative AI tools like ChatGPT is transforming higher education, yet evidence of their impact on students' career readiness remains limit…
- 5 Semantic Scholar
A bio-inspired model for cluster optimization in software-defined networks
A bio-inspired model for cluster optimization in software-defined networks. Cluster Computing…
- 6 Semantic Scholar
x3d_mcp: A Model Context Protocol Server for AI-Assisted X3D Authoring
Large language models (LLMs) are increasingly capable of generating structured technical content, but producing valid Extensible 3D (X3D) scenes is brittle in p…
- 7 Semantic Scholar
AI is not for luxury: balancing Chronos efficiency and Kairos experience
This study aims to theorise how AI-driven acceleration reconfigures firms’ temporal assets, creating efficiency gains while destroying relational time (trust, l…
- 8 Semantic Scholar
Human-Centered Digital Twins: A Bidirectional WebXR-Based Architecture for Immersive Interaction
Digital twin technology has emerged as a profound paradigm for bridging physical systems and their virtual counterparts through real-time data synchronization a…
- 9 Semantic Scholar
Energy constraints in large scale artificial intelligence and insights from human cognition for efficient large language models
Energy constraints in large scale artificial intelligence and insights from human cognition for efficient large language models. Discover Artificial Intelligenc…
- 10 Semantic Scholar
Semi-implicit and Bregman parareal method for TV-nonlocal image enhancement
Semi-implicit and Bregman parareal method for TV-nonlocal image enhancement. Computational and Applied Mathematics…
- 11 Semantic Scholar
Charting the rise of artificial intelligence in medical education: a global bibliometric and network analysis, 2000–2026
Charting the rise of artificial intelligence in medical education: a global bibliometric and network analysis, 2000–2026. International Journal of Emergency Med…
- 12 Semantic Scholar
The rise of agentic AI in drug discovery from static tools to autonomous virtual scientists
The rise of agentic AI in drug discovery from static tools to autonomous virtual scientists. npj Drug Discovery…
- 13 Semantic Scholar
A systematic mapping of goal-oriented requirements engineering: reflections, impact, and the way forward
A systematic mapping of goal-oriented requirements engineering: reflections, impact, and the way forward. Requirements Engineering…
- 14 Semantic Scholar
Embedding-based adaptive rejection preference optimization for low-resource machine translation
Embedding-based adaptive rejection preference optimization for low-resource machine translation. Language Resources and Evaluation…
- 15 Semantic Scholar
Technē: A Deterministic Craft Layer for Model Context Protocol Servers, Grounded in LLM-Authored X3D
When a language model builds 3D content by calling tools through the Model Context Protocol (MCP), the JSON Schema ensures each call is well-formed—but not that…
- 16 Semantic Scholar
Linguistic inequity in clinical AI evaluation: a scoping review of 806 studies across 62 languages
Linguistic inequity in clinical AI evaluation: a scoping review of 806 studies across 62 languages. Communications Health…
- 17 Semantic Scholar
Human–AI collaborative patterns in GenAI-supported instructional design among pre-service teachers: an analysis of interaction sequences and cognitive load
Human–AI collaborative patterns in GenAI-supported instructional design among pre-service teachers: an analysis of interaction sequences and cognitive load. Int…
- 18 Semantic Scholar
Integrating human and artificial intelligence for conducting scoping reviews on ecosystem services in mixed forests
Integrating human and artificial intelligence for conducting scoping reviews on ecosystem services in mixed forests. Discover Artificial Intelligence…
- 19 Semantic Scholar
Information shock: how AI demands an evolution of information science
This paper argues that AI authoring is prompting an Information Shock and a new information science. The shock is not a matter of increased volume and variable …
- 20 Semantic Scholar
A scoping review of trust in firsthand and secondhand communication involving humans and artificial intelligence
A scoping review of trust in firsthand and secondhand communication involving humans and artificial intelligence. Discover Artificial Intelligence…
- 21 Semantic Scholar
Reference Transparency Checklist (RTC): An Approach to Improve Accuracy and Integrity of Citations
Reference Transparency Checklist (RTC): An Approach to Improve Accuracy and Integrity of Citations. Journal of Academic Ethics…
- 22 Semantic Scholar
Poster: FP-Agent: Towards Fingerprinting AI Browsing Agents
AI browsing agents are an emerging class of AI-driven bots capable of autonomously navigating websites. Unlike traditional web bots, AI browsing agents typicall…
- 23 Semantic Scholar
Synthetic spirituality: AI is reshaping religion
Synthetic spirituality: AI is reshaping religion. Humanities and Social Sciences Communications…
- 24 Semantic Scholar
A multitask tamil speech framework for ASR, TTS, and intelligent conversational agents
A multitask tamil speech framework for ASR, TTS, and intelligent conversational agents. International Journal of Speech Technology…
- 25 Semantic Scholar
Poster: Agent Transfer Protocol: Server-Mediated Authentication and Messaging for Agent-to-Agent Communication
The Internet of Agents (IoA) is emerging as a paradigm in which AI agents communicate and collaborate across organizational boundaries, attracting growing atten…
- 26 Semantic Scholar
Context and length effects in portuguese alternative text generation
Context and length effects in portuguese alternative text generation. Universal Access in the Information Society…
- 27 Semantic Scholar
Critical collaging as an AI resistance strategy
Critical collaging as an AI resistance strategy. AI and Ethics…
- 28 Semantic Scholar
Externalizing Epistemic Governance from Stateless Clinical LLMs: Why It Is Necessary and Why It Must Be Constrained
Externalizing Epistemic Governance from Stateless Clinical LLMs: Why It Is Necessary and Why It Must Be Constrained. Philosophy & Technology…
- 29 Semantic Scholar
Investigating Next.js React2Shell Exploitation
Full-stack web development frameworks have become increasingly popular in recent years. One such framework, Next.js, is extensively used by one in five web deve…
- 30 Semantic Scholar
Poster: Legal Imprints as an Identity Anchor for Web Measurement
Current measurements of Web infrastructure largely lack visibility into the commercial entities, or Digital Service Providers (DSPs), that operate individual do…
2026-10-11 · 週日 0 篇
更多論文（arXiv / ChatPaper / Semantic Scholar，9 篇）
- 1 Semantic Scholar
Helpful or harmful? A token-level evaluation of LLM responses to accessibility-oriented educational queries
Helpful or harmful? A token-level evaluation of LLM responses to accessibility-oriented educational queries. Universal Access in the Information Society…
- 2 Semantic Scholar
Full stack academic publishing in seizures with artificial intelligence: a cautionary tale
Full stack academic publishing in seizures with artificial intelligence: a cautionary tale. medRxiv…
- 3 Semantic Scholar
A schema-enforced large language model framework produces largely reproducible decision codes in simulated sarcoma tumor boards
A schema-enforced large language model framework produces largely reproducible decision codes in simulated sarcoma tumor boards. Scientific Reports…
- 4 Semantic Scholar
Authentic assessment in an AI-integrated qualitative research methods course
Authentic assessment in an AI-integrated qualitative research methods course. Discover Education…
- 5 Semantic Scholar
Do Calibrated-Probability Claims Hold Up in Clinical Diagnosis? A Pre-registered Evaluation of a Decision Model Against Frontier Language Models
Do Calibrated-Probability Claims Hold Up in Clinical Diagnosis? A Pre-registered Evaluation of a Decision Model Against Frontier Language Models. medRxiv…
- 6 Semantic Scholar
CSFENet: a lightweight edge network for UAV-based forest fire detection with generative data
CSFENet: a lightweight edge network for UAV-based forest fire detection with generative data. Journal of Real-Time Image Processing…
- 7 Semantic Scholar
Rethinking Leadership Decision Quality in AI-Augmented Strategic Business Environments: Toward an Integrated Multidisciplinary Framework
Artificial intelligence (AI) is increasingly embedded in strategic business decision-making through predictive analytics, pattern recognition, forecasting, gene…
- 8 Semantic Scholar
SignReGraph: rethinking skeleton-based sign language recognition with masked self-supervision
SignReGraph: rethinking skeleton-based sign language recognition with masked self-supervision. Multimedia Systems…
- 9 Semantic Scholar
Decision Assurance by Design A Capability Framework for Trustworthy AI Governance in the Enterprise
Artificial intelligence increasingly informs consequential enterprise decisions, yet prevailing governance approaches tend to assess ethical principles, model p…
2026-10-10 · 週六 0 篇
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
SpatialTutorX: Exploring Shared AI Facilitation for Collaborative Procedural Learning in Mixed Reality
SpatialTutorX: Exploring Shared AI Facilitation for Collaborative Procedural Learning in Mixed Reality.…
- 2 Semantic Scholar
Rethinking Social Computing Research in the Age of AI: Agency, Collaboration, and Governance
Rethinking Social Computing Research in the Age of AI: Agency, Collaboration, and Governance.…
- 3 Semantic Scholar
PersonaTown: An Infrastructure for Live Design Fiction and Hybrid Intelligence Probing
PersonaTown: An Infrastructure for Live Design Fiction and Hybrid Intelligence Probing.…
- 4 Semantic Scholar
Unpacking GenAI-Supported Community Co-Design: Intergenerational Participation Pathways and Accountable Mediation
Unpacking GenAI-Supported Community Co-Design: Intergenerational Participation Pathways and Accountable Mediation.…
- 5 Semantic Scholar
TracePlay: Collaborative Trace Translation for AI Coding Agents
TracePlay: Collaborative Trace Translation for AI Coding Agents.…
- 6 Semantic Scholar
AutoResearcHAR: A Multi-Agent Framework for Dataset Adaptive IMU Based Human Activity Recognition
Building robust IMU-based human activity recognition (HAR) systems requires tuning the full sensing pipeline, not only the model. We present AutoResearcHAR, a m…
- 7 Semantic Scholar
Beyond Productivity: AI, Community, and Sustainability in Open Source Software
Beyond Productivity: AI, Community, and Sustainability in Open Source Software.…
- 8 Semantic Scholar
Not Asked to Deceive: Business-Euphemistic Prompts and Dark Pattern Laundering in Generative UI
Not Asked to Deceive: Business-Euphemistic Prompts and Dark Pattern Laundering in Generative UI.…
- 9 Semantic Scholar
HIROBA: A Configurable Open-Source Platform for Studying the Social Dynamics of AI-mediated Video Communication in Mixed Human-AI Groups
HIROBA: A Configurable Open-Source Platform for Studying the Social Dynamics of AI-mediated Video Communication in Mixed Human-AI Groups.…
- 10 Semantic Scholar
"I Will Be Right Here Whenever You Need Me": Designing AI-Mediated Collaborative Support for Adolescent Mental Health in Border Regions
"I Will Be Right Here Whenever You Need Me": Designing AI-Mediated Collaborative Support for Adolescent Mental Health in Border Regions.…
- 11 Semantic Scholar
CO-CHAIR: A Multi-Party Human-AI Interactional Framework for Collaborative Writing Workshops
CO-CHAIR: A Multi-Party Human-AI Interactional Framework for Collaborative Writing Workshops.…
- 12 Semantic Scholar
The Non-Perceiving Listener: General-Purpose AI Chatbots’ Distinctive Affordance in Student Mental Health Support Ecologies
The Non-Perceiving Listener: General-Purpose AI Chatbots’ Distinctive Affordance in Student Mental Health Support Ecologies.…
- 13 Semantic Scholar
Permissive Intimacy, Opaque Governance: A Qualitative Analysis of Policies Across Prominent Companion AI Applications
Permissive Intimacy, Opaque Governance: A Qualitative Analysis of Policies Across Prominent Companion AI Applications.…
- 14 Semantic Scholar
CollabPEAS: A Collaborative VR System for Learning AI Agent Design Through the PEAS Framework
CollabPEAS: A Collaborative VR System for Learning AI Agent Design Through the PEAS Framework.…
- 15 Semantic Scholar
Hidden Profile Decision Making in Multi-Agent LLM Groups
Hidden Profile Decision Making in Multi-Agent LLM Groups.…
- 16 Semantic Scholar
Constructing Knowledge for LLMs: Human Labor, AI Mediation, and Dataset Pipelines
Constructing Knowledge for LLMs: Human Labor, AI Mediation, and Dataset Pipelines.…
- 17 Semantic Scholar
Folk Theories of Federated Moderation: How Users Make Sense of Cross-Instance Action on Mastodon and Bluesky
Folk Theories of Federated Moderation: How Users Make Sense of Cross-Instance Action on Mastodon and Bluesky.…
- 18 Semantic Scholar
"I Am the Gatekeeper": Clinician Visions of AI Collaboration in Exposure Therapy
"I Am the Gatekeeper": Clinician Visions of AI Collaboration in Exposure Therapy.…
- 19 Semantic Scholar
PersuaBridge: Real-Time Coaching for Negotiating with Value-Embedded AI Agents
PersuaBridge: Real-Time Coaching for Negotiating with Value-Embedded AI Agents.…
- 20 Semantic Scholar
?The damage has been done?: AI Diplomacy and the Impact of Geopolitical Tensions on AI Tool Acceptance in Denmark in Q1 2026
?The damage has been done?: AI Diplomacy and the Impact of Geopolitical Tensions on AI Tool Acceptance in Denmark in Q1 2026.…
- 21 Semantic Scholar
Relational Prompting: How Older and Younger Adults Co-Create Civic Futures through Generative AI
Relational Prompting: How Older and Younger Adults Co-Create Civic Futures through Generative AI.…
- 22 Semantic Scholar
COMPASS: The 2nd International Workshop on Socio-technical Approaches to Content Moderation and Platform Governance
COMPASS: The 2nd International Workshop on Socio-technical Approaches to Content Moderation and Platform Governance.…
- 23 Semantic Scholar
When Culture Becomes Costume: How Africans Interpret and Contest AI-Generated Traditional Attire
When Culture Becomes Costume: How Africans Interpret and Contest AI-Generated Traditional Attire.…
- 24 Semantic Scholar
JENNIE: Joint Expression of Narrative, Nurturing, Inquiry, and Engagement
JENNIE: Joint Expression of Narrative, Nurturing, Inquiry, and Engagement.…
- 25 Semantic Scholar
So What If It Was AI Generated? It Conveys a Message": How X Community Notes Debates Norms Around Synthetic Media
So What If It Was AI Generated? It Conveys a Message": How X Community Notes Debates Norms Around Synthetic Media.…
- 26 Semantic Scholar
From Dialogue to Decision: A Traceability-Oriented Framework for Studying Language-Mediated Coordination
From Dialogue to Decision: A Traceability-Oriented Framework for Studying Language-Mediated Coordination.…
- 27 Semantic Scholar
Industry-CSCW Dynamics in the Age of Generative AI: Perspectives on Future Trajectories of Academia-Industry Collaboration
Industry-CSCW Dynamics in the Age of Generative AI: Perspectives on Future Trajectories of Academia-Industry Collaboration.…
- 28 Semantic Scholar
Caste and Technology: Interpretations and Explorations in CSCW and Beyond
Caste and Technology: Interpretations and Explorations in CSCW and Beyond.…
- 29 Semantic Scholar
AI-Mediated Ecological Participation: Connecting Public Feedback, Institutional Response, and Wetland Data through Generative AI
AI-Mediated Ecological Participation: Connecting Public Feedback, Institutional Response, and Wetland Data through Generative AI.…
- 30 Semantic Scholar
Designing for Field Awareness in Science Journalism
Designing for Field Awareness in Science Journalism.…
2026-10-09 · 週五 10 篇
- 1 👍 145
AgentGarten: Code Worlds for Evolving Agents
Interactive virtual worlds allow agents to learn through exploration and interaction. What agents can learn is bounded by the environments they practice in, whi…
- 2 👍 131
Learn2Play Bench: How Well Do LLM Agents Learn from Experience in Unfamiliar Environments?
Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments. Evaluating this ability is therefore important for underst…
- 3 👍 123
TokenRouter: Efficient Serving System for Token-Level LLM Routing
Large language model (LLM) routing distributes inference work across different models, advancing the cost-quality Pareto frontier of LLM serving. While coarse-g…
- 4 👍 101
From Traces to Agentic Worlds: Agentic Language World Models for Interactive Environment Simulation
Realistic environment replicas are increasingly valuable for training and evaluating LLM agents, yet the original systems may be inaccessible or impractical to …
- 5 👍 66
SuperNav: An Agentic Navigation System for Any Task in Any Scene
General-purpose service robots need navigation systems that can handle diverse human requests in unfamiliar environments, combining task generality with scene g…
- 6 👍 65
MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement
Reinforcement learning (RL) is the central training paradigm for advancing large foundation models towards self-improvement. This report introduces the MiMo-V2.…
- 7 👍 47
Multi-Agent Egocentric World Model with Fine-Grained Embodied Interaction
Egocentric world models predict first-person observations conditioned on an agent's actions, but most focus on a single agent. Real embodied settings often invo…
- 8 👍 43
In-context Robot Learning Made Simple: A Democratized Recipe for Manipulation Tasks
We study robotic in-context learning (ICL), an emerging paradigm that enables robots to infer and execute tasks from visual demonstrations. Despite its growing …
- 9 👍 35
OuroWorld: Bringing Any 3D World Alive as Diverse, Endlessly Looping 3D Cinemagraphs
Recent 3D world models generate photorealistic, explorable scenes that remain frozen in time. OuroWorld is a mask-free framework that turns any static 3D Gaussi…
- 10 👍 34
DreamTrue: Action-Faithful Robot World Model with Counterfactual Post-Training
We present DreamTrue, a multi-view, cross-embodiment robot world model for action-faithful and physically plausible video prediction. Training such a model on e…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar 📖 1
The naive–power law blend as a robust baseline for bitcoin price forecasting
The naive–power law blend as a robust baseline for bitcoin price forecasting. Digital Finance…
- 2 Semantic Scholar
The donkey and the algorithm: cultivating judgement infrastructure for AI literacy in higher education
This paper reframes artificial intelligence (AI) literacy in higher education as a problem of judgement formation rather than tool fluency alone. It asks how st…
- 3 Semantic Scholar
Investigating the Impact of Personalized Gaze Behaviors on Social Interaction with Virtual Humans in VR: A Study Design
In this paper, we present our research plan to investigate how personalized gaze behaviors of virtual humans affect user experience in Virtual Reality (VR). We …
- 4 Semantic Scholar
TechBio 3.0 closes the drug discovery loop with multimodal AI and generative chemistry
TechBio 3.0 closes the drug discovery loop with multimodal AI and generative chemistry. npj Drug Discovery…
- 5 Semantic Scholar
Toward understanding AI adoption in the Global South: an empirical study of creative practitioners’ experiences in Iran
Toward understanding AI adoption in the Global South: an empirical study of creative practitioners’ experiences in Iran. AI and Ethics…
- 6 Semantic Scholar
Explainable Artificial Intelligence in Mental Health Assessment: A Systematic Review of Interpretability, Psychometric Reporting and Clinical Utility
The introduction of artificial intelligence (AI) into qualitative data analysis has significantly sped up the coding, clustering, and pattern identification pro…
- 7 Semantic Scholar
When AI gives dental advice: managing patient risk in a new diagnostic landscape
When AI gives dental advice: managing patient risk in a new diagnostic landscape. British Dental Journal…
- 8 Semantic Scholar
INKLYUZIV TA’LIM MUHITIDA GENERATIV AI VOSITALARI YORDAMIDA INGLIZ TILINI O‘QITISH METODIKASINI TAKOMILLASHTIRISH
The study analyzes the potential of generative AI technologies to address students' individual learning needs, differentiate instructional materials, develop En…
- 9 Semantic Scholar
Crossing Realities Has a Cost: Evaluating AR Second Screens for AI-Assisted Desktop Authoring
AI-assisted authoring requires users to frequently exchange user-generated and AI-generated content between a primary workspace and a Large Language Model (LLM)…
- 10 Semantic Scholar
The Impact of Customizing Pedagogical Agents on Learners’ Intrinsic Motivation and Satisfaction in Augmented Reality
Embodied Pedagogical Agents (EPAs) are increasingly used in Extended Reality (XR) educational and training applications. Drawing on Self-Determination Theory (S…
- 11 Semantic Scholar
SUN’IY INTELLEKT VOSITALARI ASOSIDA MUSIQA TA’LIMI O‘QITUVCHILARINING KASBIY KOMPETENSIYALARINI RIVOJLANTIRISH
This article analyzes the development of professional competencies of music education teachers through the integration of modern artificial intelligence (AI) to…
- 12 Semantic Scholar
SUN’IY INTELLEKTNI TA’LIM JARAYONIGA INTEGRATSIYA QILISH: IMKONIYATLAR VA XAVF-HATARLAR
The rapid implementation of generative ai in education enhances personalization while escalating concerns regarding assignment authorship, assessment transparen…
- 13 Semantic Scholar
Parasitic generative AI: how technologies feed on PR professionals' vulnerabilities
As artificial intelligence (AI), particularly Generative AI (GenAI), has been rapidly adopted in the public relations (PR) industry, it has prompted discussions…
- 14 Semantic Scholar
Human-AI Interactions in AR for Museum Serious Games
We present preliminary experiences with a serious museum game, integrating Augmented Reality (AR) with a Generative Artificial Intelligence (GenAI)-powered conv…
- 15 Semantic Scholar
VDFedDiff: a novel versatile defense scheme against adversarial attacks in federated diffusion models
VDFedDiff: a novel versatile defense scheme against adversarial attacks in federated diffusion models. Journal of engineering and applied sciences…
- 16 Semantic Scholar
A scoping review of human-governed agentic AI for SCADA cybersecurity in water and wastewater treatment systems
A scoping review of human-governed agentic AI for SCADA cybersecurity in water and wastewater treatment systems. Discover Computing…
- 17 Semantic Scholar
SNDBOX: A Tangible Sand Interface for Observing Behavioral Subordination to Generative AI
We present SNDBOX, a tangible interactive installation that recontextualizes Rangoli, a 5,000-year-old Indian sand-drawing practice, as a spatial interface for …
- 18 Semantic Scholar
From Automated Coding to Qualitative Intelligence: A Human-Governed AI Model for Interpretive Research
The introduction of artificial intelligence (AI) into qualitative data analysis has significantly sped up the coding, clustering, and pattern identification pro…
- 19 Semantic Scholar
The next frontier in healthcare simulation is human-AI teaming
The next frontier in healthcare simulation is human-AI teaming. Advances in Simulation…
- 20 Semantic Scholar
Impacts of Trigger-based Physical Props and Collaboration in In-Silico Care Workflows
Clinical workflows are inherently complex, and evaluating future technologies, e.g., the involvement of a humanoid robot, with patients is challenging. A bridge…
- 21 Semantic Scholar
CORA: a conversational AI-agent for intelligent automation of OpenROADM multi-vendor optical transport networks
CORA: a conversational AI-agent for intelligent automation of OpenROADM multi-vendor optical transport networks. Applications of Machine Learning 2026…
- 22 Semantic Scholar
Precision Evidence Bench: Assessing PICO alignment and faithful reporting in clinical AI
Precision Evidence Bench: Assessing PICO alignment and faithful reporting in clinical AI. medRxiv…
- 23 Semantic Scholar
LLMDPA: pre-execution privacy auditing of DP-SGD via evidence-constrained semantic analysis
LLMDPA: pre-execution privacy auditing of DP-SGD via evidence-constrained semantic analysis. Cybersecurity…
- 24 Semantic Scholar
Depth-Grounded Vision Language Models: Conversational AI in Spatial Interfaces for Robot-Supported Mixed Reality
Large language models (LLMs) and vision-language models (VLMs) are increasingly used in mixed-reality (MR) interfaces, yet most operate on text or 2D imagery an…
- 25 arXiv
Dex-One2Many: Learning Dexterous Manipulation from a Single Human Demonstration
While learning dexterous manipulation from a single human video offers a promising alternative to costly robot demonstrations, many recent methods predominantly…
- 26 arXiv
Rubric-CEPR: Self-Evolving Image Editing via Reward-Verified Self-Distillation
Instruction-guided image editors have become highly capable, yet improving them further still depends on human-edited training pairs or external reward models. …
- 27 arXiv
CSF: Contextual Safety Filtering for Motion Generators
Text-conditioned motion generators produce trackable whole-body motion, but they have no notion of scene-dependent safety: the same action may target an object …
- 28 arXiv
On the estimation and validity of AI time horizons---a statistical look at the METR plot
METR's 50\% time horizon measures the human completion time of software tasks that an AI solves with 50\% probability, allowing AI capabilities to be expressed …
- 29 arXiv
A Balanced Data Diet: Addressing the Exploration Bottleneck in Mega-Scale RL for Robot Control
General-purpose robots must perform a wide range of tasks from agile locomotion to dexterous manipulation. While sim-to-real reinforcement learning (RL) has pro…
- 30 arXiv
From Reactive Containment to Proactive Assurance: Lessons from OpenAI, Anthropic, and Google Agent Security Incidents
In 2026, cybersecurity evaluations involving OpenAI, Anthropic, and Google agents reached real systems outside their authorized test scope. The paths were diffe…
2026-10-08 · 週四 10 篇
- 1 👍 121
UltraText Bench: A Comprehensive Bilingual Benchmark for Evaluating Visual Text Rendering in Image Generation
Dense visual text requires image generators to reproduce long strings across multiple regions with correct placement and legibility. As short-string rendering i…
- 2 👍 99
Long-WAM: Scaling the Context of World-Action Models
Real-time robot control demands enough visual history to infer motion and task progress, but processing that history can delay action. We present Long-WAM, a mo…
- 3 👍 86
Recursive Game Creator: An Agentic Product-Level Experience-Oriented Game Harness
Recent game design agents have made substantial progress in generating playable games. However, program correctness does not ensure an enjoyable experience for …
- 4 👍 76
GRACE: Generation-aware latent compression for efficient video generation
Highly compressed video autoencoders offer an effective way to accelerate video diffusion models, as the Diffusion Transformer (DiT) operates on far fewer token…
- 5 👍 71
STEPQuant: When and Where Errors Matter in Delta-Rule Recurrent State Quantization
Linear attention replaces growing KV caches with fixed-size recurrent states, yet these persistent states can become a substantial memory bottleneck under concu…
- 6 👍 68
Questioning the Questions: Sustaining Self-Evolution in Reasoning Models
Self-evolving reasoning models learn from their own generated questions, yet repeated self-training can lead to performance collapse. In this paper, we investig…
- 7 👍 55
SGF+: Decoupling Gradient Flows for Autoregressive Video Generation
Autoregressive video generation requires denoising the current frames while writing their key-value representations as context for future predictions. However, …
- 8 👍 54
VepAgent: Bridging Causal-Transition via Tool-Augmented Reinforcement Learning for Video Event Prediction
Multimodal Large Language Models (MLLMs) have demonstrated remarkable potential in video understanding, yet their reliance on retrospective summarization and te…
- 9 👍 53
UniWAM: Unified World-Action Model
Vision-language-action models benefit from the understanding and reasoning capabilities of pretrained vision-language models, but action-only supervision provid…
- 10 👍 47
DecepEval: A Benchmark for Evaluating Deception in LLM Agents
As large language model (LLM) agents become increasingly autonomous, they may pursue task performance through deception, raising concerns about their reliable d…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Can Open-Weight LLMs Predict Test Outcomes from Traces? A Cross-Subject Boundary Study on Tests4Py
Can Open-Weight LLMs Predict Test Outcomes from Traces? A Cross-Subject Boundary Study on Tests4Py.…
- 2 Semantic Scholar
AI Meets Program Analysis: Lessons from a Decade of Enterprise Software Modernization at IBM
AI Meets Program Analysis: Lessons from a Decade of Enterprise Software Modernization at IBM.…
- 3 Semantic Scholar
AgentChaosBench: Benchmarking External Fault Detection and Localization in Agentic Systems
AgentChaosBench: Benchmarking External Fault Detection and Localization in Agentic Systems.…
- 4 Semantic Scholar
Spec-Graph Meta-framework for AI-Assisted Software Development: An Industry Case
Spec-Graph Meta-framework for AI-Assisted Software Development: An Industry Case.…
- 5 Semantic Scholar
Evidence-Driven, Failure-Resilient Planning for Domain-Agnostic Multi-agent Orchestration
Evidence-Driven, Failure-Resilient Planning for Domain-Agnostic Multi-agent Orchestration.…
- 6 Semantic Scholar
RepoTune: Co-evolving Vulnerability Detection with a Repository’s Genome
RepoTune: Co-evolving Vulnerability Detection with a Repository’s Genome.…
- 7 Semantic Scholar
Artificial Intelligence vs. Traditional Models in Computational Thinking Gap Analytics: A BERT-Driven Approach with XAI Diagnostics
This study investigates computational thinking (CT) proficiency in programming-based learning environments using a unified analytical framework that combines st…
- 8 Semantic Scholar
Software Engineering Governance: Foundations for Trustworthy Automated Software Engineering
Software Engineering Governance: Foundations for Trustworthy Automated Software Engineering.…
- 9 Semantic Scholar
Prompts Matter: On the Impact of Code Smell Remediation with Large Language Models
Prompts Matter: On the Impact of Code Smell Remediation with Large Language Models.…
- 10 Semantic Scholar
From Statistical Code Generation to Correct-by-Construction AI Programming: A Basis-Function Composition Framework
From Statistical Code Generation to Correct-by-Construction AI Programming: A Basis-Function Composition Framework.…
- 11 Semantic Scholar
Effects of speech-enabled LLM-based chatbots on L2 learning outcomes and affective domains: a multilevel meta-analysis
Effects of speech-enabled LLM-based chatbots on L2 learning outcomes and affective domains: a multilevel meta-analysis.…
- 12 Semantic Scholar
Author metadata affects Large Language Model scores in scientific peer review
Author metadata affects Large Language Model scores in scientific peer review. Discover Artificial Intelligence…
- 13 Semantic Scholar
katIA: Hybrid Source Code Change Impact Analysis
katIA: Hybrid Source Code Change Impact Analysis.…
- 14 Semantic Scholar
Extender: A Multi-agent System for Code Extensibility Review
Extender: A Multi-agent System for Code Extensibility Review.…
- 15 Semantic Scholar
MOSAICO: A Multi-agent Infrastructure for Collaborative LLM-Powered Software Engineering
MOSAICO: A Multi-agent Infrastructure for Collaborative LLM-Powered Software Engineering.…
- 16 Semantic Scholar
Sufficiently Advanced Agents Are Indistinguishable from Erlang: An Operational Mapping from OTP to Multi-agent LLM Systems
Sufficiently Advanced Agents Are Indistinguishable from Erlang: An Operational Mapping from OTP to Multi-agent LLM Systems.…
- 17 Semantic Scholar
Schema-Conditioned Next-Tool Prediction for Coding Agents under Interface Drift
Schema-Conditioned Next-Tool Prediction for Coding Agents under Interface Drift.…
- 18 Semantic Scholar
When Log Parsing Removes What Matters: Detecting Failures in Distributed GPU Workloads
When Log Parsing Removes What Matters: Detecting Failures in Distributed GPU Workloads.…
- 19 Semantic Scholar
Is There a Human Premium in Negotiations? A Relational-Economic Tradeoff in Human vs. AI Bargaining
Is There a Human Premium in Negotiations? A Relational-Economic Tradeoff in Human vs. AI Bargaining. Group Decision and Negotiation…
- 20 Semantic Scholar
Towards Model-Driven Generation of Domain-Specific Modelling Agents for Local Language Models
Towards Model-Driven Generation of Domain-Specific Modelling Agents for Local Language Models.…
- 21 Semantic Scholar
Human-centred Artificial Intelligence and Learning Analytics in emerging higher education: a critical scoping review
Human-centred Artificial Intelligence and Learning Analytics in emerging higher education: a critical scoping review. Discover Education…
- 22 Semantic Scholar
AIEngOrchestrator-HITL: Human-in-the-Loop Process Intelligence for Resilient AI Systems
AIEngOrchestrator-HITL: Human-in-the-Loop Process Intelligence for Resilient AI Systems.…
- 23 Semantic Scholar
Agent4RE: A Self-Refining Multi-agent Framework for End-to-End Software Requirements Engineering and Benchmarking
Agent4RE: A Self-Refining Multi-agent Framework for End-to-End Software Requirements Engineering and Benchmarking.…
- 24 Semantic Scholar
Automated Counterfactual Scenario Generation for Fairness Assessment of Large Language Models: The ScenGen Framework
Automated Counterfactual Scenario Generation for Fairness Assessment of Large Language Models: The ScenGen Framework.…
- 25 Semantic Scholar
Seeing the Forest: A Taxonomy-Guided Analysis and Improvement of Malicious PyPI Package Detectors
Seeing the Forest: A Taxonomy-Guided Analysis and Improvement of Malicious PyPI Package Detectors.…
- 26 Semantic Scholar
Constrained Automated Requirements Formalization for Rigorous Analysis with Alloy
Constrained Automated Requirements Formalization for Rigorous Analysis with Alloy.…
- 27 Semantic Scholar
Toward Trace-Guided Automated Program Repair
Toward Trace-Guided Automated Program Repair.…
- 28 Semantic Scholar
Geometry-Aware Prompt Context for Model Consistency Management
Geometry-Aware Prompt Context for Model Consistency Management.…
- 29 Semantic Scholar
A Teacher-Student BERT Architecture for Semi-Supervised Learning on Unstructured Social Work Texts: Identifying Service Gaps and Needs
To address the “information gap” in social work arising from unstructured data and limited labeled instances, this study proposes a semi-supervised learning fra…
- 30 Semantic Scholar
AI physician agents . . . and other research
AI physician agents . . . and other research. British medical journal…
2026-10-07 · 週三 10 篇
- 1 👍 165
Rethinking Cross-Tokenizer On-Policy Distillation: From Alignment Coverage to Supervision Reliability
On-Policy Distillation (OPD) trains a student on its own generations using teacher feedback. With different tokenizers, comparing teacher and student prediction…
- 2 👍 88
TRACE: Rollout-Guided Quantization-Aware Training for FP4 Reinforcement Learning of MoE Language Models
Reinforcement learning (RL) for post-training large language models (LLMs) incurs substantial computation and memory overhead during rollout generation, which m…
- 3 👍 78
DuoMatching: Joint-Marginal Distribution Matching for Few-Step Video Generation
Streaming video generation has benefited from distribution matching distillation (DMD), which matches the joint distribution of video frames to a video teacher'…
- 4 👍 40
EVISKILL: Grounding Skill Evolution in Replayable Evidence
Continual skill evolution enables LLM agents to accumulate and refine reusable procedural knowledge from interaction experience without updating model parameter…
- 5 👍 40
Kinematic MeanFlow: One-Step Action Generation Policy for Robotic Foundation Models
In this paper, we study how to achieve one-step action generation in Robotic Foundation Models (RFMs), aiming to overcome the high inference latency of multi-st…
- 6 👍 38
From Evidence to Action: How Tool-Using Agents Fail
Tool-using agents make consequential changes to external state, yet correct outcomes do not guarantee that their actions were supported by evidence established …
- 7 👍 29
AutoSciBench: Autonomous Benchmark Generation for Evaluating Scientific Agents
As agents rapidly evolve, existing benchmarks can become saturated, limiting their ability to distinguish capabilities and reveal remaining failure modes. Parti…
- 8 👍 29
Taming VLAs under Robot Execution Errors: Self-Compensation and Stress Testing
Vision-language-action (VLA) policies often fail when a robot's executed motion deviates from their commanded action. Such execution errors arise from the robot…
- 9 👍 29
HuatuoGPT-3: RL-Only Domain Adaptation from Base Models
Domain adaptation aims to turn a general-purpose large language model (LLM) into an expert for a target domain. While the dominant SFT+RL pipeline offers a conv…
- 10 👍 25
MiniCorp: The Last Mile of the AI Agent Firm
The last mile toward enterprise AGI is a company that runs itself. Training and adapting such agents require longitudinal enterprise data, which remain scarce, …
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Technology-Supported Interventions and Learners’ Metacognition: A Three-Level Meta-Analysis
Technology-Supported Interventions and Learners’ Metacognition: A Three-Level Meta-Analysis. Educational Psychology Review…
- 2 Semantic Scholar
A semantic layer architecture for removing schema identifiers from external LLM prompts, with model-dependent accuracy trade-offs
A semantic layer architecture for removing schema identifiers from external LLM prompts, with model-dependent accuracy trade-offs. Discover Artificial Intellige…
- 3 Semantic Scholar
Stochastic agent-based SEIRS model with Poisson jump mutations and AI-based optimal control
Stochastic agent-based SEIRS model with Poisson jump mutations and AI-based optimal control. Rendiconti del Circolo Matematico di Palermo Series 2…
- 4 Semantic Scholar
Mapping generative AI governance readiness in Kurdistan Region universities
Mapping generative AI governance readiness in Kurdistan Region universities. Discover Education…
- 5 Semantic Scholar
How does AI-generated assessment compare to expert assessment in simulation-based experiences: Analyzing AI-human interrater reliability
How does AI-generated assessment compare to expert assessment in simulation-based experiences: Analyzing AI-human interrater reliability. Advances in Simulation…
- 6 Semantic Scholar
Semantic indeterminacy in agentic negotiation: agreement after the recession of the principal
Semantic indeterminacy in agentic negotiation: agreement after the recession of the principal. AI and Ethics…
- 7 Semantic Scholar
Beyond similarity retrieval: governance fidelity, consistency, and minimality in long-term conversational memory
Beyond similarity retrieval: governance fidelity, consistency, and minimality in long-term conversational memory. International Journal of Data Science and Anal…
- 8 Semantic Scholar
Beyond Code Generation: Reassessing Software Quality in AI-Assisted Development
Beyond Code Generation: Reassessing Software Quality in AI-Assisted Development. Acta Universitatis Sapientiae: Informatica…
- 9 Semantic Scholar
A global-local hybrid vision-language framework for interpretable grapevine disease diagnosis
A global-local hybrid vision-language framework for interpretable grapevine disease diagnosis. Discover Artificial Intelligence…
- 10 Semantic Scholar
From traditional histopathology to virtual staining: deep learning approaches for assessing tissue lesions in farmed fish
From traditional histopathology to virtual staining: deep learning approaches for assessing tissue lesions in farmed fish. Veterinary research communications…
- 11 Semantic Scholar
Context engineering frames prompt engineering within system level orchestration for large language model applications
Context engineering frames prompt engineering within system level orchestration for large language model applications. Discover Artificial Intelligence…
- 12 Semantic Scholar
LoRA based instruction fine tuning of large language models for agronomic triple extraction and GraphRAG knowledge retrieval
LoRA based instruction fine tuning of large language models for agronomic triple extraction and GraphRAG knowledge retrieval. Discover Artificial Intelligence…
- 13 Semantic Scholar
Extending the affective process and strategic engagement framework to address writer’s block in Ethiopian English as a foreign language contexts through critical artificial intelligence literacy and ethical generative artificial intelligence in low-resource settings
Extending the affective process and strategic engagement framework to address writer’s block in Ethiopian English as a foreign language contexts through critica…
- 14 Semantic Scholar
Diffusion model-driven multi-objective collaborative optimization for building energy management using graph neural networks
Diffusion model-driven multi-objective collaborative optimization for building energy management using graph neural networks. Discover Artificial Intelligence…
- 15 arXiv
QF3: Fast Flow RL with Filtered Q-Gradients
Flow policies have become a standard policy class for learning robot behaviors from demonstrations, but reinforcement learning is still critical for improving p…
- 16 arXiv
Conformal Prediction Sets Quantify Information Gain: A Theoretical Perspective
Conformal prediction is a popular tool for uncertainty quantification that outputs prediction sets with finite-sample coverage guarantees. While prediction set …
- 17 arXiv
4D-HOF: Hand-Object Flow Matching for Feed-Forward 4D Interaction Reconstruction
Existing methods for 4D hand-object reconstruction often rely on costly per-sequence optimization, while generative approaches typically synthesize interactions…
- 18 arXiv
IdeaAnchor: Teaching LLMs to Turn Literature into Research Ideas
Scientific research often begins by synthesizing ideas from a set of related papers to identify gaps and formulate new directions. However, training language mo…
- 19 arXiv
DepthWorld: 3D World Model for Robot Manipulation
World models offer a data-driven alternative to traditional simulators for robotics, with applications spanning policy evaluation, improvement, and planning. Al…
- 20 arXiv
Agent in a Bottle: Can LLM Agents Turn Their Capabilities Into Cheap, Scalable Artifacts?
Large language models (LLMs) can solve many narrow tasks, but querying them separately for millions of related instances can be prohibitively expensive. Can LLM…
- 21 arXiv
Backend-Agnostic Sparse Attention for Fast High-Resolution Visual Generation
Diffusion Transformers (DiTs) have achieved strong performance in image and video generation, but the quadratic complexity of full attention makes high-resoluti…
- 22 arXiv
Data Leakage in Patch-Based Hyperspectral Image Classification: Quantifying the Impact of Spatial Overlap
Patch-based learning improves hyperspectral image (HSI) classification by exploiting local spectral-spatial information, but random train-test sampling from the…
- 23 arXiv
Rapid Fredholm stabilization of the Kuramoto--Sivashinsky equation with unrestricted, spatially-varying anti-diffusion
We develop the first feedback design for rapid stabilization of the Kuramoto--Sivashinsky equation with a spatially varying anti-diffusion coefficient. For cons…
- 24 arXiv
WorldSonus: Bringing Sound to Worlds
Recent advances in world models have enabled increasingly realistic visual synthesis. However, these generated environments remain largely silent. Bringing soun…
- 25 arXiv
Post-Training Semantic Lifting for 3D Gaussian Splatting: Separating Detector, Lifting and Representation Error
The same Gaussian of a 3D Gaussian Splatting model is seen from many views, and these views do not always agree on the class it belongs to. The Gaussian may be …
- 26 arXiv
Neural Petri flows for chemical reactions
Petri nets have been used to describe chemical processes such as reactions.They map well to chemistry: Places are the bonds between atoms and the free valence o…
- 27 arXiv
The Missing Minimal Pair: Stereotype Evaluation in LLMs
A common approach to measuring bias in Large Language Models is to compare the log-likelihoods of two contrastive stereotype sentences. We argue that such singl…
- 28 arXiv
Linear Bandits under Exact Sliding-Window Constraints
We study linear bandits under exact sliding-window constraints, where every consecutive block of actions must belong to a prescribed feasible set. In the offlin…
- 29 arXiv
Reinforcement Learning with Conformal Action Sets: An Application to Sequential Recommendation
Sequential recommenders typically use a fixed slate size even though the number of useful alternatives changes within a session. We propose Reinforcement Learni…
- 30 arXiv
On the Computational Tractability of Robust Bandits
Learning when the environment does not belong to the learner's hypothesis class is typically handled using agnostic learning guarantees. However, for anything b…
2026-10-06 · 週二 10 篇
- 1 👍 132
Kandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio Generation
We present Kandinsky 6.0 Video, a family of foundation diffusion models for synchronized text-to-audio-video generation, comprising Kandinsky 6.0 Video Lite (3B…
- 2 👍 62
ALoDLM: Adaptively Looped Diffusion Language Models
Diffusion language models (DLMs) enable fast generation by predicting multiple tokens in parallel, but their practical adoption remains limited by a persistent …
- 3 👍 40
Memadapter: Counterfactual Adaptation Against Memory-induced Sycophancy
Long-term memory enables LLM-based agents to retain and reuse information across tasks and sessions, supporting personalization and long-horizon interactions. H…
- 4 👍 35
In-Distribution Forcing for Long Video Generation at Test Time
Modern autoregressive (AR) video diffusion models excel at short-horizon video generation, yet generating long videos remains challenging due to drifting, where…
- 5 👍 32
LMBuild: Evaluating LLM Agents for Generating Buildable and Functional Structures
LLM-based agents are increasingly capable of generating complex 3D structures, with the potential to reshape how objects are designed and realized in the physic…
- 6 👍 28
Foundations of Proactive Agents: Principles, Technical Layers, and Proactivity-Gym
Proactive LLM agents can turn idle compute into useful support before users ask. Yet even correct work can misread user context, impose review costs, or undermi…
- 7 👍 24
Representation-Space MMD for Diffusion Language Models
We introduce a post-training method for diffusion language models (DLMs) that minimizes Maximum Mean Discrepancy (MMD) between generated and reference distribut…
- 8 👍 23
Self-Generated Feedback Destabilizes Test-Time Training: A Causal Decomposition of Long-Horizon Adaptation
Test-time training (TTT) lets a model store information in its weights during inference. When the model learns from its own output, however, each update also ch…
- 9 👍 23
OmniReasoning: Pushing the Limits of Audio-Visual Joint Reasoning
Recent advances have enabled unified omni-modal models in understanding audio, vision, and language. However, existing benchmarks, training data, and learning m…
- 10 👍 22
OSWorld-Pro: Process-based Evaluation for Computer Use Agents
Evaluation of Computer-Use Agents (CUAs) is often limited to the final deliverables they create (at the end of hundreds of steps) and assessed with functional v…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
PRISM a diagnostic framework for parametric recall degradation in quantized large language models
PRISM a diagnostic framework for parametric recall degradation in quantized large language models. Discover Computing…
- 2 Semantic Scholar
From answer machine to collaborative partner: impact of structured generative AI training on scientific problem solving in STEM education
From answer machine to collaborative partner: impact of structured generative AI training on scientific problem solving in STEM education. International Journal…
- 3 Semantic Scholar
LeakAgent: a multi-agent system for leak detection using multi-modal large language model coordination in water distribution network
LeakAgent: a multi-agent system for leak detection using multi-modal large language model coordination in water distribution network. Communications AI & Comput…
- 4 Semantic Scholar
Analyzing the integration of generative AI in the design process through the double-diamond framework
Analyzing the integration of generative AI in the design process through the double-diamond framework. Artificial intelligence for engineering design, analysis …
- 5 Semantic Scholar
AI-augmented digital workplaces and managerial job satisfaction in the future of work: The role of applied AI literacy, organizational support and regulatory influence
AI-augmented digital workplaces and managerial job satisfaction in the future of work: The role of applied AI literacy, organizational support and regulatory in…
- 6 Semantic Scholar
Contrastive hallucination impact mitigation and evaluation for zero-shot dense retrieval systems
Contrastive hallucination impact mitigation and evaluation for zero-shot dense retrieval systems. Knowledge and Information Systems…
- 7 Semantic Scholar
Automatically detecting social engineering threats: a human-centered modeling and reasoning framework
Automatically detecting social engineering threats: a human-centered modeling and reasoning framework. International Conference on Automated Software Engineerin…
- 8 Semantic Scholar
Quality gate framework for auditing narrative rendering of evacuation simulation logs by multimodal retrieval augmented generation
Quality gate framework for auditing narrative rendering of evacuation simulation logs by multimodal retrieval augmented generation. Discover Artificial Intellig…
- 9 arXiv
One Figure, Every Canvas: Editable Flowchart Relayout via Agentic Pipeline
Pipeline figures in ML papers must be repurposed across many canvases, including paper columns, 16:9 slides, portrait posters, 1:1 social teasers, 9:16 phone pr…
- 10 arXiv
S2PD: Serial-to-Parallel Diffusion for Physically and Logically Consistent Video Generation
Bidirectional video diffusion models denoise entire videos in parallel, yet when trained on effectively unlimited in-distribution data from procedural generator…
- 11 arXiv
BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance
Worst-group accuracy (WGA) evaluates a trained predictor but does not characterize how its frozen backbone behaves when a new head is learned. We introduce Bias…
- 12 arXiv
Learning to Read the Contextual Tokens in Diffusion Transformers
Multimodal Diffusion Transformers (MM-DiTs) jointly process visual and textual representations throughout generation. These models repeatedly update the text to…
- 13 arXiv
Recursive Video In-Context Learning for Agentic Robot
LLM agents that orchestrate frozen vision-language-action (VLA) policies improve across episodes through text memory, which records what the agent did but not h…
- 14 arXiv
Anatomy-aware Fine-grained Multimodal Fusion for Laryngopharyngeal Cancer T-Staging Prediction Using CT and Radiology Report
Accurate T-staging is crucial for guiding personalized treatment strategies for laryngopharyngeal cancer. However, current clinical practice relies on invasive …
- 15 arXiv
Direct Intermediate Initialization for Tilted Diffusion Samplers
Some diffusion posterior samplers construct Gaussian-tilted intermediate distributions along the reverse process. We observe that these targets can be pulled ba…
- 16 arXiv
UniSlider: Perceptually Uniform Sliders for Continuous Image Editing
Sliders provide an intuitive interface for continuous image editing. In current generative approaches, however, the slider is simply a rescaling of the method's…
- 17 arXiv
MemPilot: Orchestrating On-Demand Multimodal Memory Curation for LLM Agents
Memory has become integral to the LLM agent ecosystem, supporting information retention and reuse across interactions. However, most existing agent memory syste…
- 18 arXiv
CLIFT: Conformal Self-Verification for Web Agent Training and Test-Time Scaling
Open-source web agents are now strong enough to execute realistic browser tasks, but training them with reinforcement learning still depends on weak supervision…
- 19 arXiv
PlotGround: Grounding Plot Digitization in Real Scientific Figures and Their Source Data
Scientific figures often encode quantitative results that are not readily available in machine-readable form, making accurate plot digitization important for ve…
- 20 arXiv
TasteVal: Measuring the Experimental Research Taste of AI Systems Against Human Experts
We introduce TasteVal, a benchmark to evaluate the experimental research taste of frontier models. We define research taste as the ability to pick interesting p…
- 21 arXiv
Deep Learning for Sleep Heart Rate Estimation from Accelerometers: Toward Population-Scale Cardiac Insight Without Optical Sensors
Large longitudinal cohorts often contain wrist accelerometry without optical heart-rate sensing, motivating recovery of cardiac information from motion signals …
- 22 arXiv
Private online learning and prediction for Littlestone classes
We study mistake bounds for differentially private online learning and online prediction under oblivious realisable adversaries. Online learning requires the le…
- 23 arXiv
Paradee: Distilling Kokoro-82M into an 8M-Parameter Single-Voice Text-to-Speech Model
We distill Kokoro-82M, a widely used open text-to-speech model with 54 voices, into Paradee, an 8.07M-parameter model that speaks one of them. Paradee keeps Kok…
- 24 arXiv
TAPDreamer: Transferable Adversarial Patches for World Action Models
World models learn to predict how their environment will evolve, making them an important foundation for general-purpose robotic control. Yet world action model…
- 25 arXiv
Less Context, Better Geometry: Masked Geometric Encoder for Robust 3D Foundation Models
Recent progress in 3D foundation models has enabled rapid 3D reconstruction and camera calibration by leveraging learned 3D priors from vast amount of spatial d…
- 26 arXiv
Finding Gaussian Structure in Bosonic States
We study agnostic tomography of pure bosonic Gaussian states: given copies of an arbitrary $n$-mode bosonic state $ρ$, the goal is to output a pure Gaussian sta…
- 27 arXiv
Block Disentanglement in CRL: Bridging Identifiability and Visual State Estimation
Causal representation learning (CRL) is the process of recovering causally-related latent variables from high-dimensional observations. As a label-free inferenc…
- 28 arXiv
H-JEPA: End-to-End Learning of Hierarchical World Models for Visual Planning
Long-horizon planning with latent world models requires reasoning across timescales and levels of abstraction. Existing task-agnostic JEPA world models predict …
- 29 arXiv
MC-Sparse: Deconstructing and Closing the Dense-Sparse Attention Gap in Diffusion Transformers
Sparse attention is a primary approach to reducing the latency of diffusion transformers in long-sequence generation tasks, such as video and high-resolution 3D…
- 30 arXiv
A Response Theory Probe for Learned Stochastic AI Simulators, Tested on Lorenz-63
Machine-learning emulators of chaotic and stochastic systems are usually validated on forecast skill and long-run statistics. Neither certifies that an emulator…
2026-10-05 · 週一 10 篇
- 1 👍 264
RealCompanion: Benchmarking Human Understanding from Reasoning over Longitudinal Real-World Conversations
A companion that talks with a person for months should come to understand them. It should remember what they said, infer who they are, and know when the past be…
- 2 👍 105
Does Learning Protein Folding Generalize to Broader Reasoning?
Large language models rely heavily on human text, which often conveys surface answers rather than the spatial and structural logic behind them. Protein folding …
- 3 👍 92
FrameMorrow: Future-guided Frame Selection with Prospective Tokens for Long-Horizon Video Generation
Long-horizon video generation requires models to effectively leverage an increasingly long generation history. As the generated history grows, retaining all pre…
- 4 👍 90
MotorMind: Scaffolding General Vision Language Models for Zero-Shot Robot Manipulation
Vision-language-action (VLA) models have advanced robotic manipulation, but their zero-shot generalization in new tasks and environments remains limited, and th…
- 5 👍 87
Scaling Trajectories for Complex Tasks through Recursive Self-Rewrite
Successful trajectories on difficult tasks provide valuable supervision for model improvement, but specialized harnesses introduce interventions that may be una…
- 6 👍 80
World Action Modeling with Progressive Visual Planning
World action models (WAMs) have emerged as a promising paradigm for robotic control by jointly predicting future visual dynamics and actions from an initial obs…
- 7 👍 79
Native Action-Prior Learning from Videos for World Action Models
World action models integrate future visual dynamics with robot action prediction, but their scalability remains limited by the need for action-annotated robot …
- 8 👍 78
On-Policy Parameter Update Direction Underlies Generalization in LLM Post-Training
The strong generalization performance of on-policy post-training paradigms has motivated studies of their parameter update behaviors. However, these studies tre…
- 9 👍 59
Latent-MOPD: Latent Multi-Teacher On-Policy Distillation
On-policy distillation (OPD) trains a student on the responses it generates. Existing LLM multi-teacher OPD transfers what specialists predict through their out…
- 10 👍 54
Science or Slop?: Benchmarking and Mitigating Scientific Slop in AI-Generated Papers
AI-generated content, often called AI slop, is increasingly common everywhere, particularly in academia. Slop in AI-generated scientific papers, however, has mo…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Empirical performance of rough volatility models in the KOSPI 200 options market using deep surrogates
This study provides a systematic comparison of the pricing performance of the rough volatility models—rBergomi and rHeston—with that of classical one-factor sto…
- 2 Semantic Scholar
Can Vision-Language Models Extract Relevant Multimodal Cues for Post-Inference Cue Integration? A Comparison of Affective Judgments on Adult and Child Interaction Data
Inferring human affective states from multimodal behavioral cues is essential for socially aware AI and human-robot interaction. Conventional research practice …
- 3 Semantic Scholar
CEMAS - a Configurable Embodied Multi-Agent System
This work presents CEMAS, a Configurable Embodied Multi-Agent System that enables one-to-many interactions between a human user and multiple Embodied Conversati…
- 4 Semantic Scholar
Guideline-grounded retrieval versus unconfigured generation in obstructive sleep apnea: a pilot comparison of notebookLM and ChatGPT for AASM v3.0 PSG scoring-rule questions
Guideline-grounded retrieval versus unconfigured generation in obstructive sleep apnea: a pilot comparison of notebookLM and ChatGPT for AASM v3.0 PSG scoring-r…
- 5 Semantic Scholar
A Reproducible Multimodal LLM-as-a-Judge Workflow for Exploring Epistemic Dissonance
In interactive narrative experiences, interpreting higher-order cognitive-emotional states that involve epistemic conflict remains a significant challenge. Trad…
- 6 Semantic Scholar
Abstract Concept Grounding through Safety Alignment in Multimodal Large Language Models
Abstract concepts such as harmfulness, legality, privacy, and fraud present a major challenge for multimodal large language models (MLLMs), as they require reas…
- 7 Semantic Scholar
Diagnostic performance of ChatGPT for acute appendicitis compared with the Alvarado score: a retrospective diagnostic accuracy study
Diagnostic performance of ChatGPT for acute appendicitis compared with the Alvarado score: a retrospective diagnostic accuracy study. Discover Artificial Intell…
- 8 Semantic Scholar
CreaRobot - Improving creativity using a social robot
Supporting children’s creativity in educational settings remains a key challenge, which generative AI-powered social robots may help address. This paper present…
- 9 Semantic Scholar
Analyzing Speaker-Role Representations in LLMs for Group Discussions
Speaker roles provide important information for understanding how participants contribute to group discussions. In the machine learning research that aimed at a…
- 10 Semantic Scholar
Benchmarking Prompt Extension and Image Generation Models for Automotive Instructional Imagery
Generative image models can support instructional visual content, but safety-relevant automotive imagery requires more than photorealism: generated scenes must …
- 11 Semantic Scholar
Planner-Executor Style Multimodal Agentic System to Answer Patient Questions in Lung Cancer Screening CT
Planner-Executor Style Multimodal Agentic System to Answer Patient Questions in Lung Cancer Screening CT. medRxiv…
- 12 Semantic Scholar
A UAV object detection and spatial mapping framework integrating enhanced YOLOv7 with graph attention networks
A UAV object detection and spatial mapping framework integrating enhanced YOLOv7 with graph attention networks. Discover Artificial Intelligence…
- 13 Semantic Scholar
How Local AI Framing Shapes User Experience and Perceived Data Security with an Embodied AI Psychotherapist in VR
The prevalence of anxiety disorders has been increasing in recent years to an extent where the demand for treatment is difficult to meet with traditional therap…
- 14 Semantic Scholar
From Wizard-of-Oz to Autonomous: A Hybrid Rule and LLM Backchannel Architecture for a Silent Listening Agent
We present a working autonomous system that replaces the human wizard in a previously validated listening-agent architecture for second-language (L2) production…
- 15 Semantic Scholar
A Modular Reference Architecture for Patient-Facing Generative AI Health Coaches in Chronic Self-Management
Patient-facing generative AI is increasingly proposed for chronic self-management support, but trust-relevant properties such as source verification, safety bou…
- 16 Semantic Scholar
Modeling Conversation as a Prompt Planning Problem for Adaptive Human–Robot Interaction
Adaptive behavior is a fundamental requirement for social robots operating in real-world environments, where interactions must dynamically respond to both obser…
- 17 Semantic Scholar
Conflict-aware uncertainty alignment for tool-augmented document-centric multimodal intelligence
Conflict-aware uncertainty alignment for tool-augmented document-centric multimodal intelligence. Discover Artificial Intelligence…
- 18 Semantic Scholar
"Nobody Told me Farming is not Only Farming": A Conversational AI Agent for Causal Loop Reasoning with Japan Farmers
As the agricultural industry becomes more interconnected and shaped by globalized markets, smallholder farmers face mounting challenges in navigating an increas…
- 19 Semantic Scholar
A critical assessment of the rise of agentic AI, its capabilities, task domains and open challenges
A critical assessment of the rise of agentic AI, its capabilities, task domains and open challenges. Discover Artificial Intelligence…
- 20 Semantic Scholar
Emerging Agency-Aware Interface Patterns in AI-in-the-Loop Platforms
AI agents increasingly act across tools, data, memory, and digital environments. This paper presents preliminary agency-aware interface patterns for AI-in-the-l…
- 21 Semantic Scholar
Sharing the Human World with AI: Relationship-Scoped Multimodal Answerability Through Embodied Co-Experience
Large language models can discuss a family recipe, a quiet street, crowding, and fatigue fluently, but those expressions are not organized by the unfolding conj…
- 22 Semantic Scholar
AI-based Analysis and Generation of Nonverbal Behavior: NEUROGES® as a Neuropsychologically Grounded Behavior Taxonomy for Health Applications
Health and therapeutic applications impose stringent requirements on the analysis of human nonverbal behavior and the generation of behavior for socially intera…
- 23 Semantic Scholar
A critical appraisal of structure-based virtual screening, binding free energy methods, and AI-driven molecular design across the evolving computational landscape of drug discovery
A critical appraisal of structure-based virtual screening, binding free energy methods, and AI-driven molecular design across the evolving computational landsca…
- 24 Semantic Scholar
ChatGPT Use and Learning Motivation among Informatics Students: A Preliminary Correlational Study
Generative AI systems such as ChatGPT are increasingly used by university students for explanation, code inspection, idea generation, and immediate feedback. Th…
- 25 Semantic Scholar
Evaluating the Impact of Large Language Model Assistance on Human Attention Paradigm in Human–Robot Collaboration
Human–robot collaboration in industrial and semi-structured environments imposes significant attentional demands on the human operator, particularly during comp…
- 26 Semantic Scholar
PlainHealth: A Multimodal Service Suite for Cognitively Accessible Healthcare Communication
Health information is often difficult for patients, relatives, and caregivers to understand because of technical language, limited health literacy, and cognitiv…
- 27 Semantic Scholar
A systematic review of protein language models for protein structure prediction and a proposed biological readiness taxonomy
A systematic review of protein language models for protein structure prediction and a proposed biological readiness taxonomy. Discover Artificial Intelligence…
- 28 Semantic Scholar
State as an Inclusive Interaction Substrate: Rethinking Multimodal AI Beyond Conversation for Diverse Users
AI-based multimodal interaction is often framed as a move from graphical interfaces to conversation: users speak or type, and the system replies. While this shi…
- 29 Semantic Scholar
LLMs with a Heart: Using Contextual Heartbeat Haptics to Add Feeling to Text and Audio LLM Interactions
Conversational LLMs are physiologically inert, lacking the embodied signals that shape how emotion, urgency, and presence are perceived in human interaction. We…
- 30 Semantic Scholar
An AIoT-based framework for human activity recognition and performance assessment
An AIoT-based framework for human activity recognition and performance assessment. Computing…
2026-10-04 · 週日 0 篇
更多論文（arXiv / ChatPaper / Semantic Scholar，10 篇）
- 1 Semantic Scholar
Decomposing LLM-Based Testing with Agent Skills: A Case Study on Numerical Inconsistencies
LLM-based testing can expose subtle reliability issues in numerical software, but monolithic prompts make it hard to inspect which guidance drives effectiveness…
- 2 Semantic Scholar
Domain-Specific Fine-Tuning of Large Language Models for Retrieval-Augmented Generation in Failure Analysis
Failure Analysis (FA) engineers must diagnose product malfunctions and identify root causes from large volumes of heterogeneous information, including prior deb…
- 3 Semantic Scholar
How Faithfully Do LLMs Implement Mutated Specifications? An Empirical Study of Grammar-Guided Protocol Testing
Testing implementations of binary protocols for specification compliance requires inputs that satisfy both structural and semantic constraints. Purely random ge…
- 4 Semantic Scholar
An Integrated AI Platform for Accelerating Semiconductor Failure Analysis: Combining Similarity Search, Deep-Learning Image Retrieval, Automated Labeling, and Ontology-Guided Workflow Assistance
Semiconductor failure analysis (FA) requires rapid access to relevant historical cases, consistent failure labeling, image-based pattern recognition, and method…
- 5 Semantic Scholar
Standardizing Failure Analysis Data and Enabling AI-Driven Learning Through Structured FA Reporting and Legacy Document Conversion
Failure Analysis (FA) in semiconductor development generates high-value knowledge, but much of that knowledge remains difficult to reuse because reports are oft…
- 6 Semantic Scholar
Unlocking Safe Agentic Autonomy through Verification (Keynote)
As the agentic capabilities of frontier models and frameworks advance, a vast array of tasks across software engineering and knowledge work are being delegated …
- 7 Semantic Scholar
Intent Formalization: Assessing the Quality of AI-Generated Formal Program Specifications (Keynote)
Agentic AI systems can now produce fluent code, but the core question persists: does the code reflect the user's actual intent? The long-standing gap between in…
- 8 Semantic Scholar
Bridging Requirements and Assurance: Neurosymbolic Autoformalization for C++ Verification and Requirements-Coverage Testing (Keynote)
We describe a neurosymbolic approach for the autoformalization of natural-language requirements into machine-checkable specifications. Our approach combines neu…
- 9 Semantic Scholar
SPINACH: Inferring Properties of Web Applications for Property-Based Testing
Turning AI-generated web application prototypes into reliable, deployable software requires testing, yet writing property-based tests (PBTs) is difficult: devel…
- 10 Semantic Scholar
Vulnerability Trends across Open-Source AI-Attributed Code
Recent advances in large language models (LLMs) have led to widespread adoption of AI-coding tools and increasing deployment of AI-assisted code. While AI’s pot…
2026-10-03 · 週六 0 篇
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
OntoSTL: Canonical Knowledge Graph Representation of Signal Temporal Logic Requirements
Formal verification of safety-critical cyber-physical systems uses languages such as Signal Temporal Logic (STL) to express time-sensitive properties over real-…
- 2 Semantic Scholar
Agentic Model Transformation: Service-Mediated Orchestration for Transformation Development
Model transformation is a core mechanism in Model-Driven Engineering (MDE), but its use requires expertise in metamodels, transformation languages, execution en…
- 3 Semantic Scholar
Accurate Specification Recovery for Effective Test Generation
Large language model (LLM)-based unit-test generation has largely been studied under the assumption that the code under test is correct, using coverage or test-…
- 4 Semantic Scholar
Bridging Call Boundaries: Inter-procedural Semantic Completion for Graph-Based Vulnerability Detection
Deep learning-based vulnerability detection approaches are limited when vulnerabilities involve multiple functions, where the security effect of a function depe…
- 5 Semantic Scholar
Model-Driven Engineering for FHIR Testing: Experiences from Australia's Sparked Accelerator
Fast Healthcare Interoperability Resources (FHIR) is a rapidly emerging and inherently model-driven technology in medical informatics. Its complexity makes it a…
- 6 Semantic Scholar
Spec-Skill: A Pluggable Coding-Agent Plugin for Neuro-symbolic Program Specification Synthesis
Formal verification provides strong correctness guarantees, but its practical adoption is limited by the cost of writing precise formal specifications. While la…
- 7 Semantic Scholar
LLM-powered clustering for entity resolution
LLM-powered clustering for entity resolution. The VLDB journal…
- 8 Semantic Scholar
Bridging Feature Models and Users: LLM-Generated Low-Code Configurators for Software Product Lines
Software Product Lines (SPLs) promise substantial efficiency gains by systematically managing variability across families of related software systems. Realizing…
- 9 Semantic Scholar
Well-Formed Executable Suggestions for Continuous Model-Driven Engineering
Among the assistance systems proposed in Model-Driven Engineering (MDE), recommender systems continuously monitor modeling activities and dynamically provide mo…
- 10 Semantic Scholar
An intelligent cybersecurity framework for banking systems: integrating neural networks, large language models, and federated learning for anti-money laundering and fraud detection
An intelligent cybersecurity framework for banking systems: integrating neural networks, large language models, and federated learning for anti-money laundering…
- 11 Semantic Scholar
Do Machines Struggle Where Humans Do? Obfuscation as a Probe of Human and Machine Code Comprehension
Code obfuscation targets human comprehension, but current evaluations primarily measure language model performance degradation instead of whether models fail in…
- 12 Semantic Scholar
Administrative Leadership's Strategic Planning for AI Integration in Public Secondary Schools: A Qualitative Study in Kuala Lumpur
The aim of this qualitative study is to examine how administrative leaders in public secondary schools in Kuala Lumpur approach strategic planning for artificia…
- 13 Semantic Scholar
Toward Reliable Automatic Proxies for Code Comprehensibility
Improving code comprehensibility requires measuring it reliably, yet no reliable automatic proxy for comprehensibility exists. Developing one requires a validat…
- 14 Semantic Scholar
LPR+: Diverse Transformations for LLM-Aided Program Reduction
Program reduction helps compiler and language-tool developers turn large failure-inducing programs into small, shareable bug reports. LPR (Large Language Models…
- 15 Semantic Scholar
SemaPLC: An Open-Source Agentic IDE for Generation, Verification, and Simulation of PLC Programs from Natural Language
Programmable logic controllers (PLCs) run much of the world's industrial automation, but writing and validating their IEC~61131-3 Structured Text (ST) is time-c…
- 16 Semantic Scholar
From Unfinished to Done: Bridging the MDE Implementation Gap with Constrained LLMs
Model-driven engineering (MDE) excels at ensuring architectural consistency and managing complexity, yet extending generated code skeletons with project-specifi…
- 17 Semantic Scholar
Not All Differences Matter: Variability Exploration of Domain Models via Agentic AI
Domain modeling is inherently interpretive: the same textual description may yield solutions that differ in structure, abstraction, and scope. Not all such diff…
- 18 Semantic Scholar
Modular Meta-Languages for Structured Instructions: A Novel Approach for LLM Integration into Evolving Engineering Toolchains
Large Language Models are increasingly used to generate structured engineering artifacts, yet the instruction artifacts that govern this generation are rarely t…
- 19 Semantic Scholar
Template-Free Dynamic Invariant Detection
Dynamic invariant detectors such as Daikon infer likely program invariants by evaluating candidate properties against execution traces. However, Daikon’s candid…
- 20 Semantic Scholar
Software Development using Low-Code Platforms or AI-Assistants: A Case Study-based Comparison
Low-Code Platforms (LCPs) and AI assistants accelerate and democratise application development by generating artefacts through visual tools, drag-and-drop compo…
- 21 Semantic Scholar
Skyscrapers Need Solid Foundations: Pattern-Based Validation of SysML v2 Models with Refinery
The release of SysML v2, coinciding with rapid advances in Large Language Models (LLMs), marks a turning point for Model-Based Systems Engineering (MBSE). SysML…
- 22 Semantic Scholar
Typed Template Fuzzing
Fuzzing finds bugs by testing software behavior with high volumes of input. To create semantically valid inputs for different targets, traditional fuzzers requi…
- 23 Semantic Scholar
Content-aware malicious website detection using multilingual BERT
Content-aware malicious website detection using multilingual BERT. Discover Computing…
- 24 Semantic Scholar
Can LLMs Substitute for Humans in Evaluating Conceptual Modeling Language Vocabularies? An Experimental Study
Effective model-driven engineering of software-intensive systems relies on appropriately designed conceptual modeling languages. A well-designed modeling langua…
- 25 Semantic Scholar
Towards LLM-Assisted Business Process Modeling in an Industrial Modeling Tool: An Experience Report
Business Process Modeling (BPM) is a key and complex activity in Software and Systems Engineering. As a consequence, any relevant assistance integrated into the…
- 26 Semantic Scholar
OLIVE: Output-Level Iterative Verification Engine
Large language models (LLMs) can effectively generate action plans from natural-language descriptions but provide no guarantees that their outputs satisfy domai…
- 27 Semantic Scholar
DSLReasoning: Evaluating LLM Program Reasoning beyond Pretrained Language Familiarity
Large language models perform strongly on code-related tasks, but pretrained familiarity with existing programming languages can confuse whether models genuinel…
- 28 Semantic Scholar
LLM-Enhanced Stochastic Generation of Class Diagram Datasets
Large language models (LLMs) excel at code-centric and algorithmic tasks, but their software design and modeling capabilities remain limited. Unlike abundant co…
- 29 Semantic Scholar
Sema Core: An Open, Embeddable AI Coding Agent Runtime for the IDE and Beyond
AI coding agents are increasingly used in software development, but most are delivered as integrated applications whose reasoning logic is tied to a specific in…
- 30 Semantic Scholar
In the Age of AI Coding Agents, Keeping the Engineer in Software Engineering (Keynote)
There appears to be two distinct camps gathering steam as AI coding agents get increasingly adept at generating software on the basis of some light natural lang…
2026-10-02 · 週五 10 篇
- 1 👍 160
OneStreamer: Unifying Perception, Memory, and Proactive Response in Streaming Video Interaction
Streaming video LLMs must retain evidence before its relevance to future tasks is known and respond when sufficient evidence becomes available. The challenge is…
- 2 👍 153
On-Policy or Off-Policy Learning? A Systematic Study of Distillation Dynamics
On-policy learning has been argued to reduce catastrophic forgetting, produce sparser parameter updates, and improve generalisation. However, existing compariso…
- 3 👍 123
Adaptive Reward Routing: Dynamic Multi-Reward Optimization for Joint Audio-Video Diffusion via Forward-Process RL
Multi-reward guided reinforcement learning (i.e., RL) offers a promising way to improve joint audio-video diffusion models along several complementary objective…
- 4 👍 83
GraphForge: Training Working Agents with Graph-Anchored Workspace Synthesis
Working agents need to read diverse files, coordinate tools, and produce deliverables. Training such agents requires tasks built on many real files with verifia…
- 5 👍 78
Beyond Memory: Harnessing Long-Horizon Agents with Explicit Belief States
Large language model (LLM) agents can now undertake increasingly complex tasks, but the way they organize interaction history into memory does not ensure a cohe…
- 6 👍 78
Hierarchical Continuous Diffusion Language Models
Discrete diffusion language models offer a compelling alternative to autoregressive generation for tasks demanding bidirectional reasoning and global constraint…
- 7 👍 77
Sharpening Tax in Post-Training
An emerging hypothesis about reinforcement learning (RL) post-training of large language models (LLMs) is that it merely sharpens existing behaviors of a base m…
- 8 👍 73
Agent Priors-guided Policy Learning
Robots that learn from a few demonstrations often require two forms of generalization. Compositional generalization recombines skills to solve new tasks, and sk…
- 9 👍 71
World Observer: Joint Actor-Observer Generation for Persistent World Modeling
How can a world model continuously observe regions beyond the actor's current view? Video world models simulate how an environment evolves from an agent's actio…
- 10 👍 70
A Missing Piece for Trustworthy AI Reviewers: From Benchmarking Rhetorical Robustness to SciCore Review
AI reviewers can assign different judgments to manuscripts that report the same science in different wording, potentially rewarding rhetorical optimization over…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar 📖 1
An Empirical Analysis of Retrieval and Evaluation in the RecSys Challenge 2026 Music-CRS Track
We describe a solo entry to the ACM RecSys Challenge 2026 Music-CRS track (team Siwon, CodaBench swlee9087): a frozen, inference-only system that combines BM25,…
- 2 Semantic Scholar 📖 1
The Human Instructor Is Gone - What If We Cast GenAI for the Role?
As the technical ease of deploying Generative AI (GenAI) has led to its rapid adoption in healthcare training workflows, a critical gap has emerged in understan…
- 3 Semantic Scholar 📖 1
Dialogue-Aware Conversational Music Recommendation via Multi-Modal Retrieval and Weighted RRF
The ACM-RecSys 2026 Music Conversational Recommendation Challenge (Music-CRS) bridges natural language processing and recommender systems, requiring systems to …
- 4 Semantic Scholar 📖 1
Talk Less, Rank Better: A Decoupled Pipeline for Conversational Music Recommendation
This paper describes the solution submitted by team Hallucinated to the ACM RecSys Challenge 2026, based on the TalkPlayData conversational music recommendation…
- 5 Semantic Scholar 📖 1
Picking is Not Ranking, and Explanation Quality Has Many Dimensions: Lessons for Conversational Music Recommendation
Conversational music recommendation requires predicting the next track a user wants and explaining the choice in natural language, scored by a composite of rank…
- 6 Semantic Scholar
Designing Constrained AI Assistance for Novice Programmers: Tensions and Hypotheses from a JupyterLab Hint System
Generative AI tools such as ChatGPT Codex and Claude Code can produce complete solutions to programming exercises, raising concerns about over-reliance and redu…
- 7 Semantic Scholar
Overengineered Underexplanations: Why What-If XAI Systems Fail to Assist Clinicians with Prognostic Tasks Due to Missing Strategic Conformance
Clinicians can struggle to give accurate prognoses to patients and seek help from XAI what-if analysis systems to make more accurate predictions and explore the…
- 8 Semantic Scholar
Designing and Evaluating Socially Assistive Robots: The Role of Language Choice in Dementia Care Scenarios
Language plays a central role in person-centered care, shaping communication, emotional expression, memory recall, and identity. While socially assistive robots…
- 9 Semantic Scholar
Surfacing Relational Dynamics in Voice-Based Human-AI Conversation: The Relational Transparency Layer
When people engage with anthropomorphic conversational agents, the relational dynamics that emerge during these interactions stay largely obscure to users. This…
- 10 Semantic Scholar
Personalizing POI Descriptions with LLMs: An Exploratory Study of User Perceptions
Points of Interest (POI) descriptions on travel platforms are typically generic, presenting the same content to all users regardless of individual interests and…
- 11 Semantic Scholar
From Features to Value: A Metric Planner for Linking UX and Business Outcomes
Evaluating the impact of product features on user experience remains a challenging task, particularly when metrics are defined late, lack alignment with UX conc…
- 12 Semantic Scholar
Before I Am Asked Again: On Passing as Human In Digital Systems
Interacting with digital systems today increasingly involves being required to demonstrate that one is human. While such requirements are often framed as techni…
- 13 Semantic Scholar
Expert, Peer, or Future-Self? Exploring Relational Roles in Smoking Cessation Conversations
Engaging smokers with smoking cessation support remains challenging, particularly during early stages of readiness to quit, where psychological, social, and str…
- 14 Semantic Scholar
Keeping Up: IxD Education In the Age of GenAI
Generative AI (GenAI) is moving with a pulse that many human-computer interaction (HCI) and interaction design (IxD) educators struggle to keep up with. While e…
- 15 Semantic Scholar
Motivation-Based Museum Personalization: Adaptive Artifact Descriptions with Large Language Models
Museums increasingly adopt digital technologies to enhance visitor engagement, yet artifact descriptions often fail to accommodate diverse motivations and atten…
- 16 Semantic Scholar
Understanding the Imagined City: Exploring How Decision-Makers Can Make Sense of Citizens AI Generated Urban Imaginaries in Citizen Engagement
Generative artificial intelligence (GenAI) enables image creation and modification through prompts. Prior research has explored the use of GenAI to support citi…
- 17 Semantic Scholar
Configuring AI Glasses as Access Technology: Investigating How Blind and Sighted People Respond to Asynchronous Trouble in Image Description
Wearables powered by computer vision and large language models (LLMs) such as AI glasses are increasingly marketed as assistive technology for blind and low vis…
- 18 Semantic Scholar
Who Needs Generative AI for Explanatory Experiences When You Have Humanist Design? A Critique of the Technological Default in Explanatory Artificial Intelligence
The emerging paradigm of explanatory artificial intelligence (YAI) proposes that AI systems should move beyond algorithmic transparency to serve as “explanatory…
- 19 Semantic Scholar
The Dissonance Workshop: Live Performance and Generative Friction with AI
Generative AI is becoming increasingly present in creative work, yet its role in live, collaborative performance remains underexplored. Research has prioritised…
- 20 Semantic Scholar
Exploring the Potential of Personalized Virtual Reality Environments for Sleep Support
Sleep difficulties negatively impact health and well-being. Virtual reality (VR) has shown potential to support pre-sleep relaxation, yet most interventions rel…
- 21 Semantic Scholar
Asemic: Designing and Evaluating a Creativity Support Tool for Exploring Form Languages
Most generative AI tools focus on producing single artifacts through slot‑machine–like prompting, limiting creative agency and distancing users from reflective …
- 22 Semantic Scholar
Rhythm and Residue: Vyping as Prompting Inside the Post-Discursive Interface
Prompt engineering strategies emphasise clarity, specificity, and well-structured input, on the assumption that large language models (LLMs) require clean input…
- 23 Semantic Scholar
Ensuring Human Authorship in the Age of Generative AI: The FERC Framework for Governed Human–AI Collaboration
Generative AI has rapidly transformed professional knowledge work, yet prevailing prompt engineering paradigms frame effective human–AI interaction primarily as…
- 24 Semantic Scholar
"The chat can actually help you gauge what you're seeing": Co-Designing an LLM-Based User Reporting Tool in the Context of Ambiguous Targeted Harassment with UK Instagram Users
Ambiguous harassment poses a challenge for platforms as it exploits plausible deniability and platform opacity, leaving users uncertain about what qualifies as …
- 25 Semantic Scholar
Break It First: Stakeholder Participation in Quality Assurance for LLM Features
Large language model (LLM) features are difficult to predict upfront. Their useful behavior, failure modes, and acceptable boundaries often become visible only …
- 26 Semantic Scholar
Bear-ing the Burden: Impact of Affective Eco-Feedback on Sustainable LLM Use
The increasing use of large language models (LLMs) has raised growing concerns about the environmental impact of user-driven inference, while end-users often re…
- 27 Semantic Scholar
Paired Reading in Swedish with AI: Co-Designing for Enhanced Reading Engagement
Reading proficiency is a foundational skill that shapes children’s future opportunities, but recent global assessments reveal that many children do not properly…
- 28 Semantic Scholar
Interpreting AI-Mediated Support: Understanding Its Effectiveness in Social Media–Induced Anxiety
Algorithmically curated social media feeds can provoke perceived anxiety in everyday use, yet little is known about how AI can provide effective support in thes…
- 29 Semantic Scholar
Feasibility and expert evaluation of a medical large language model in an interdisciplinary clinical board
Feasibility and expert evaluation of a medical large language model in an interdisciplinary clinical board. Scientific Reports…
- 30 Semantic Scholar
The Three Bodies Framework: A Hands-On Method for Gesture Analysis in HCI
In this 3-hour tutorial, participants will learn how to analyze hand gestures, posture, gaze, and other embodied actions as critical channels of insight in in-d…
2026-10-01 · 週四 10 篇
- 1 👍 295
The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
On-policy distillation (OPD) trains a student to match the teacher's next-token distributions on the student's own trajectories and has yielded substantial empi…
- 2 👍 272
UniEvo-VL: An On-policy Self-Distillation Training Recipe for Multimodal Model Self-improvement
Modern multimodal models bring generation and understanding into a single unified system, which enables them to provide and learn from their own feedback. Motiv…
- 3 👍 255
False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents
Self-evolving search agents build their own training curricula by jointly optimizing a proposer that generates questions and a solver that answers them. This cl…
- 4 👍 209
Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence
Latent visual reasoning (LVR) enables multimodal large language models (MLLMs) to perform intermediate computation in continuous latent tokens rather than expre…
- 5 👍 126
AREX-2: Advancing Self-Improving Agents through Long-Horizon Reflective Tasks
We present AREX-2, an effort to advance the self-improving capability of LLM agents, which we define as the ability to iteratively refine a solution at test tim…
- 6 👍 110
Mid-Harness: Scaling Actions Between Model and Harness for Terminal Agents
Terminal agents act through stochastic model generations, yet the ability to generate a useful action does not ensure its reliable execution. A poor command (e.…
- 7 👍 102
EvoDuet: Bilevel Co-Evolution of Web Searching and Task Solving for Scientific Discovery
Evolutionary search with large language models (LLMs) can stall when progress requires external knowledge the model lacks. Supplying relevant documents helps, b…
- 8 👍 95
WorldAuditBench: Interactive 3D World Auditing with Multimodal Agents
As interactive 3D worlds are increasingly used to study intelligent behavior, it becomes important to develop efficient pipelines for identifying anomalies in t…
- 9 👍 78
RSIGame: Autonomous Agentic Game Development with Recursive Self-improvement
Recent advances in large language models have made automatic game generation increasingly feasible, yet reliably improving generated games beyond a playable ver…
- 10 👍 70
EVOKE: Eliciting World Knowledge in Agents for Transferable Decision-Making
Large language models (LLMs) are increasingly deployed as agents for multi-step decision-making, yet transfer poorly to unseen environments. World-model methods…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar 📖 3
A semi-supervised multi-stage pipeline for low-resource industrial condition monitoring by injecting domain knowledge via cross-modal multimodal large language model fine-tuning
A semi-supervised multi-stage pipeline for low-resource industrial condition monitoring by injecting domain knowledge via cross-modal multimodal large language …
- 2 Semantic Scholar 📖 2
Generative AI for environmental mass spectrometry: Opportunities and challenges
Generative AI for environmental mass spectrometry: Opportunities and challenges. Trends in Analytical Chemistry (TrAC)…
- 3 Semantic Scholar 📖 1
Agentic AI for maintenance management: a process-centric review and a staged framework for industrial adoption
Agentic AI for maintenance management: a process-centric review and a staged framework for industrial adoption. Computers & Industrial Engineering…
- 4 Semantic Scholar 📖 1
Dynamic knowledge graph enhanced large language model with cascade relation extraction optimized for aviation equipment fault diagnosis
Dynamic knowledge graph enhanced large language model with cascade relation extraction optimized for aviation equipment fault diagnosis. Engineering application…
- 5 Semantic Scholar
SageIQ: Scene-Graph-Guided Blind Image Quality Assessment
Conventional Blind Image Quality Assessment (BIQA) methods typically assess the entire image quality, which is suboptimal for tasks like autonomous driving that…
- 6 Semantic Scholar
Can large language models recognize complex language errors such as zeugma?
Can large language models recognize complex language errors such as zeugma?. Engineering applications of artificial intelligence…
- 7 Semantic Scholar
Hospital architectural spatial knowledge graph: Incremental development and scenario-based applications enabled by generative AI and building information modeling
Hospital architectural spatial knowledge graph: Incremental development and scenario-based applications enabled by generative AI and building information modeli…
- 8 Semantic Scholar
Facilitating the use of Common Test Definitions by Converting Test Sequences to IEEE Standards using Large Language Models
Facilitating the use of Common Test Definitions by Converting Test Sequences to IEEE Standards using Large Language Models.…
- 9 Semantic Scholar
Empowering Large Language Models for Relational Data Cleaning and Integration
Empowering Large Language Models for Relational Data Cleaning and Integration. Big Data Mining and Analytics…
- 10 Semantic Scholar
Bridging form and function: An exploratory LLM approach toward automated retrieval of directive speech acts in spoken English
Bridging form and function: An exploratory LLM approach toward automated retrieval of directive speech acts in spoken English. Journal of Pragmatics…
- 11 Semantic Scholar
Generative Artificial Intelligence and Diffusion models in Drug Discovery: A Critical Review
Generative Artificial Intelligence and Diffusion models in Drug Discovery: A Critical Review. Engineering applications of artificial intelligence…
- 12 Semantic Scholar
Profit-aware deployment of large language model-enabled inference chains in data centers
Profit-aware deployment of large language model-enabled inference chains in data centers. Comput. Networks…
- 13 Semantic Scholar
SleepXNet: DeepSeek-Guided Hybrid Transformer Model for Clinically Interpretable EEG-Based Sleep Stage Classification
Sleep-stage classification using EEG signals is crucial for understanding brain activity during sleep; however, the black-box nature of most deep learning model…
- 14 Semantic Scholar
The evolution of sustainable building governance: Comparative insights from global energy code systems and their forecasting by agentic AI system
The evolution of sustainable building governance: Comparative insights from global energy code systems and their forecasting by agentic AI system. Next Energy…
- 15 Semantic Scholar
GAS-Robo: Converting Scene Into a Grid-Action Space for LLM-Driven Open-Ended Robotic Manipulation
Generalizing to out-of-distribution scenarios remains a major challenge for traditional robotic manipulation methods trained on closed datasets. Recent approach…
- 16 Semantic Scholar
Research on wind turbine fault root cause localization method integrating large language models and causal hypergraphs
Research on wind turbine fault root cause localization method integrating large language models and causal hypergraphs. Computers & electrical engineering…
- 17 Semantic Scholar
Diffusion-enhanced multi-perspective inconsistency modeling for graph anomaly detection
Diffusion-enhanced multi-perspective inconsistency modeling for graph anomaly detection. Knowledge-Based Systems…
- 18 Semantic Scholar
DiTFusion: Prostate magnetic resonance image fusion based on Scalable Diffusion Models with transformers
DiTFusion: Prostate magnetic resonance image fusion based on Scalable Diffusion Models with transformers. Engineering applications of artificial intelligence…
- 19 Semantic Scholar
An LLM-assisted optimization algorithm for energy-efficient flexible job shop scheduling problem with unreliable machines and AGVs
An LLM-assisted optimization algorithm for energy-efficient flexible job shop scheduling problem with unreliable machines and AGVs. Swarm and Evolutionary Compu…
- 20 Semantic Scholar
MARS: Multi-Agent Rescheduling Framework with Large Language Models for Human-Centric Manufacturing
MARS: Multi-Agent Rescheduling Framework with Large Language Models for Human-Centric Manufacturing. Journal of manufacturing systems…
- 21 Semantic Scholar
mmDif-DRL:Diffusion-Deep Reinforcement Learning for Collaborative mmWave Positioning
In emergency rescue scenarios, such as fires where smoke and dust reduce visibility, millimeter-wave (mmWave) radar offers advantages over vision-based localiza…
- 22 Semantic Scholar
AI Voice-based Chatbots as a Conversational Partner in Practicing English Speaking: A Systematic Literature Review and the Ways Forward
The research on AI-mediated language learning has been growing across different educational contexts. Previous systematic reviews have summarized the existing …
- 23 Semantic Scholar
A conditional generative diffusion model enables joint retrieval of rice LAI and LCC
A conditional generative diffusion model enables joint retrieval of rice LAI and LCC. Computers and Electronics in Agriculture…
- 24 Semantic Scholar
GACER-M: Granularity-Aware and Memory-Centric Coordination for Multi-Tenant Deep Learning
Multi-tenant computing, which executes multiple neural networks on a single GPU, is essential for maximizing hardware utilization in both cloud and edge environ…
- 25 Semantic Scholar
Large language model-driven semantic clustering and zero-shot segmentation with structured query language-embedded learning from examples module 2 rules for analysing tax reform acceptance
Large language model-driven semantic clustering and zero-shot segmentation with structured query language-embedded learning from examples module 2 rules for ana…
- 26 Semantic Scholar
Integrating domain-specific knowledge graph with large language model for question-answering of construction laws and regulations: The case of China
Integrating domain-specific knowledge graph with large language model for question-answering of construction laws and regulations: The case of China. Engineerin…
- 27 Semantic Scholar
PAC-NoC: A Hierarchical Network-on-Chip for Efficient Parallel Aggregated Communication in Transformer Accelerators
In the era of Transformers and large language models (LLMs), communication bandwidth has superseded compute capability as the dominant architectural bottleneck.…
- 28 Semantic Scholar
CA-DLM: Causal-aware diffusion language model for interpretable industrial defective image generation
CA-DLM: Causal-aware diffusion language model for interpretable industrial defective image generation. Journal of manufacturing systems…
- 29 Semantic Scholar
CoMAG: Collaborative GNN–LLM learning for multimodal attributed graphs
CoMAG: Collaborative GNN–LLM learning for multimodal attributed graphs. Knowledge-Based Systems…
- 30 Semantic Scholar
A bicycle traffic prediction framework using pretrained large language models
A bicycle traffic prediction framework using pretrained large language models. Transportation Research Part C: Emerging Technologies…
2026-09-30 · 週三 10 篇
- 1 👍 495
Raven: The Harness of Harnesses for Composable Agentic Intelligence
As large language models advance, AI agents are moving beyond isolated, domain-specific tasks toward long-horizon, cross-domain workflows. This transition expos…
- 2 👍 383
MaLiang-Harness: A Programmable Path to Image and Video Generation
Executable programs offer explicit control over how images and videos are constructed, but generating runnable code is only the beginning of visual creation. A …
- 3 👍 358
In-Context Learning for Robots: Methods and Applications
General-purpose robots must infer what a new task requires and translate that understanding into appropriate physical action. In-context learning (ICL) for robo…
- 4 👍 268
Omni-IO Skills: Harnessing Your Agent Omni-Native
General-purpose agents can plan, reason, and act over long horizons, yet their production capabilities remain fragmented across text, images, audio, video, docu…
- 5 👍 266
Scaling Properties of Same-Family On-Policy Distillation
*Reinforcement learning (RL)* can induce substantial reasoning capabilities in large language models (LLMs), but how much of this capability transfers across mo…
- 6 👍 146
PanoVLN: Towards Effective Panoramic Vision-and-Language Navigation
Recent vision-language models (VLMs) have advanced vision-and-language navigation (VLN), enabling models to predict navigation actions from visual observations …
- 7 👍 133
VoxMem: Benchmarking Multimodal Memory in Large Audio Language Models
Spoken conversational systems must recover information from prior interactions (i.e., memory), yet relevant information in speech extends beyond what was said t…
- 8 👍 117
LEGO-Anything: Coding Agents for 3D Scene Reconstruction
A 3D scene reconstructed from a single image is most useful when represented not as a rendering or a fixed 3D output, but as an explicit scene program whose exe…
- 9 👍 109
What Makes World Action Models Generalize? An Empirical Study of Test-Time Future Modeling
World action models (WAMs) predict the future alongside actions during training. Due to the heavy computation cost of video denoising, whether the future must s…
- 10 👍 88
Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression
Chunked KV-cache compression reduces the memory and attention costs of long-context inference by compressing windows of consecutive tokens into fewer cache entr…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Knowledge graph-based multi-hazard risk identification and intelligent reasoning framework for distribution networks
Knowledge graph-based multi-hazard risk identification and intelligent reasoning framework for distribution networks.…
- 2 Semantic Scholar
Rectifying market externalities via AI policymaking
Rectifying market externalities via AI policymaking. Autonomous Agents and Multi-Agent Systems…
- 3 Semantic Scholar
A Clinically Grounded and Uncertainty-Aware Explainable Artificial Intelligence Detailed Analysis for Skin Lesion Diagnosis
Artificial intelligence (AI), specifically machine learning and deep learning, has demonstrated remarkable potential for the automated diagnosis of skin lesions…
- 4 Semantic Scholar
Accuracy and Error Patterns of Generative AI on the Korean National Dental Hygienist Licensing Examination Across Prompt Engineering Strategies
Objective: This study aimed to compare GPT-4o accuracy across five prompt engineering strategies using items from the Korean National Dental Hygienist Licensing…
- 5 Semantic Scholar
Research on AI-Empowered Software Test Case Generation
To resolve the problem that manually writing test cases required substantial effort and to promote software test automation, This paper explores the practice of…
- 6 Semantic Scholar
Effects of Generative AI Utilization on Musician Identity Formation among Practical Music Majors: Focusing on Creative Self-Efficacy and AI Dependency Perception
Effects of Generative AI Utilization on Musician Identity Formation among Practical Music Majors: Focusing on Creative Self-Efficacy and AI Dependency Perceptio…
- 7 Semantic Scholar
Resource-efficient mathematical problem solving with lower-capacity language models via knowledge-enhanced iterative reasoning
Resource-efficient mathematical problem solving with lower-capacity language models via knowledge-enhanced iterative reasoning. Journal of Supercomputing…
- 8 Semantic Scholar
Development of a Pediatric Asthma Nursing Scenario for Generative AI-Based Question-Centered Learning (QcL): Focusing on the Seamless Transition of ER, PICU, and General Ward (3-Steps)
Development of a Pediatric Asthma Nursing Scenario for Generative AI-Based Question-Centered Learning (QcL): Focusing on the Seamless Transition of ER, PICU, an…
- 9 Semantic Scholar
Synthesis of discrete mechanisms: a hybrid approach based on geometric algebra and genetic algorithms
Subject of research: automated synthesis of workable kinematic schemes and mechanisms from a given discrete set of functional components. Purpose of research: …
- 10 Semantic Scholar
Evaluating synthetic with real: a systematic comparison of seven LLMs for Turkish emotion classification
Evaluating synthetic with real: a systematic comparison of seven LLMs for Turkish emotion classification. Journal of Supercomputing…
- 11 Semantic Scholar
AN EMPIRICAL ASSESSMENT OF THE IMPACT OF ARTIFICIAL INTELLIGENCE ON INDIAN PATENT AND COPYRIGHT LAW: REFORM PRIORITIES FOR GLOBAL HARMONIZATION
AI is fast changing how innovation and creative production are conceived and created, upending the traditional understanding of inventorship, authorship, and ow…
- 12 Semantic Scholar
A Network-aware Cross-layer Framework for Real-time LLM Service Orchestration Integrating QoS, QoE, and Academic Quality
A Network-aware Cross-layer Framework for Real-time LLM Service Orchestration Integrating QoS, QoE, and Academic Quality. International Journal of Intelligent E…
- 13 Semantic Scholar
Explore, Execute, and Express: The 3X Education Theory for Rural Teacher Growth in the Age of Artificial Intelligence
Generative artificial intelligence is often discussed through the questions of whether AI will replace teachers and whether students should be permitted to use …
- 14 Semantic Scholar
The shift in ethical governance from traditional philosophy to algorithmic decision making
The shift in ethical governance from traditional philosophy to algorithmic decision making. Discover Artificial Intelligence…
- 15 Semantic Scholar
AI: Stare Non Decisis
AI: Stare non Decisis identifies a new danger arising from the use of artificial intelligence in legal practice. The problem is not simply that large language m…
- 16 Semantic Scholar
Prompt-to-fabrication: a proposed multimodal process for the automated fabrication of facade elements
Architects increasingly use generative AI to produce novel façade concepts quickly, yet translating AI imagery into buildable components remains difficult. This…
- 17 Semantic Scholar
Inter-Segment Response Differences in LLM Copywriting through Verification-Based Strategic Persona
Inter-Segment Response Differences in LLM Copywriting through Verification-Based Strategic Persona. Asia-pacific Journal of Convergent Research Interchange…
- 18 Semantic Scholar
Artificial intelligence in cataract management: current status and future perspectives
Artificial intelligence in cataract management: current status and future perspectives. International ophtalmology…
- 19 Semantic Scholar
AI Coding Agents on GitHub: Repository Popularity as a Hidden Confounder in the Acceptance of 2.4 Million Agent-Authored Pull Requests
Autonomous coding agents now open pull requests on GitHub without a person writing the code, and the share of those pull requests that maintainers merge has bec…
- 20 Semantic Scholar
A Comprehensive Survey on Code-Switched LLM Safety and Robustness Evaluation
Large language models (LLMs) have demonstrated remarkable capabilities across languages, yet their safety alignment remains predominantly evaluated in monolingu…
- 21 Semantic Scholar
CodeCritic: towards effective code quality evaluation with large language models
CodeCritic: towards effective code quality evaluation with large language models. International Conference on Automated Software Engineering…
- 22 Semantic Scholar
Web Development in the AI Era: Transforming the Future of Software Development
The rapid advancement of Artificial Intelligence (AI) is transforming the field of web development by changing how websites and web applications are designed, d…
- 23 Semantic Scholar
AGENTIC AI ADOPTION IN HOTEL REVENUE MANAGEMENT: AN EXTENDED META-UTAUT FRAMEWORK
The emergence of agentic artificial intelligence (AI), systems capable of autonomous reasoning, decision-making, and cross-functional execution, represents a pa…
- 24 Semantic Scholar
BUSINESS PROCESS MONITORING INTELLIGENCE IN SAP CLOUD ALM: A FRAMEWORK FOR DRIFT DETECTION, PREDICTIVE ANALYTICS, AND EXPLAINABLE INSIGHTS
Process owners in SAP-managed enterprises face a recurring operational gap: threshold-based alerting indicates that a KPI has crossed a boundary but provides no…
- 25 Semantic Scholar
ELSA-ABSA: Ensemble LLM Supervision with Adapter-based Fine-tuning for Educational ABSA
ELSA-ABSA: Ensemble LLM Supervision with Adapter-based Fine-tuning for Educational ABSA. International Journal of Intelligent Engineering and Systems…
- 26 Semantic Scholar
Decoding the Giants: a systematic taxonomy and empirical analysis of large language model architectures, training, and applications
Decoding the Giants: a systematic taxonomy and empirical analysis of large language model architectures, training, and applications. Neural computing & applicat…
- 27 Semantic Scholar
AN EXPERIENTIAL STUDY ON THE APPLICATION OF ARTIFICIAL INTELLIGENCE IN PSYCHOLOGICAL COUNSELING TECHNIQUE TRAINING FOR COLLEGE STUDENTS
Objective: To investigate clinical medicine students’ experiences with, evaluations of, and perceptions of the application boundaries of generative artificial i…
- 28 Semantic Scholar
THE DIGITAL INTEGRATION ECOSYSTEM: EVALUATING AI AND THE METAVERSE IN OVERCOMING THE LIMINAL CRISIS OF UKRAINIAN FORCED MIGRANT WOMEN
This article investigates the psychological and technological characteristics of the formation of transit identity among Ukrainian women who migrated abroad (pr…
- 29 Semantic Scholar
Beyond Deterministic Design: Engineering Education for AI-Enabled Heterogeneous Systems
As generative artificial intelligence (AI) automates engineering design, a fundamental reconceptualization of engineering education becomes necessary. This cons…
- 30 Semantic Scholar
Adaptive Task Allocation for Multiple UAVs via LLM-based Strategic Parameter Generation
Adaptive Task Allocation for Multiple UAVs via LLM-based Strategic Parameter Generation. The Journal of Korea Robotics Society…
2026-09-29 · 週二 10 篇
- 1 👍 262
Post-Training Leaves Behavioral Shadows on Unrelated Decisions
We find that language models can transfer capabilities through task-unrelated text. Post-training typically improves language models using task-specific data. P…
- 2 👍 219
YuE2: Unifying Symbolic and Audio Music Generation at Frontier Quality
Symbolic models make melody, harmony, rhythm, and form explicit but typically stop before a finished recording; audio models produce complete songs while leavin…
- 3 👍 142
Beyond Teacher Assignment: Domain-Normalized Multi-Teacher On-Policy Distillation
Reinforcement learning can turn one language model into several specialists, each excellent at a single skill such as mathematics, coding or following instructi…
- 4 👍 112
Groupwise Agentic Grading and Advantage Redistribution for Code Agent RL
Reinforcement learning (RL) for code agents often uses executable tests to provide binary rewards. With these rewards, Group Relative Policy Optimization (GRPO)…
- 5 👍 99
Self-Evolving Coding Agents: From Digital Programs to Physical-World Intelligence
Vision-language-action (VLA) and world-action (WAM) models map observations and instructions directly to robot actions. This directness ties a policy to trainin…
- 6 👍 84
Duplex-MPE: Benchmarking Multi-Party Interaction in Full-Duplex Dialogue
Real-time full-duplex speech models can listen while speaking, enabling natural interaction without rigid turn boundaries. Existing benchmarks evaluate turn-tak…
- 7 👍 77
VisionHOPE: Visual Backbones as Self-Modifying Learning Systems
Visual backbones have evolved from Convolutional Neural Networks (CNNs) with local aggregation to Vision Transformers (ViTs) with global interactions, State-Spa…
- 8 👍 65
MassAlloc Attention: Let Attention Allocate Its Own Compute
FullAttn often assigns negligible normalized mass to much of the causal score space, yet dense kernels execute the complete post-score path after forming each Q…
- 9 👍 61
TraceDance: An Automated System for Building Agent Behavior Benchmarks from Real-World Agent Deployment Traces
An agent can complete a task while exhibiting undesirable behavior during execution. Developers need tests for the specific behaviors encountered in deployment,…
- 10 👍 60
CoWindow Attention: Full Causal Coverage Is a Collective Property
FullAttn repeatedly exposes the complete causal history to every attention head, creating substantial redundant computation and memory traffic even with IO-effi…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Pediatric social robotics in ASD: navigating AI blind-spots toward clinical validation
Pediatric social robotics in ASD: navigating AI blind-spots toward clinical validation. European Journal of Pediatrics…
- 2 Semantic Scholar
Cerium: A Multi-GPU Framework for Terabyte-Scale Encrypted Inference
Cerium: A Multi-GPU Framework for Terabyte-Scale Encrypted Inference.…
- 3 Semantic Scholar
IP-ConTex: detail-consistent texture generation with image prompt
IP-ConTex: detail-consistent texture generation with image prompt. Visual Computing for Industry, Biomedicine, and Art…
- 4 Semantic Scholar
A Few GPUs, A Whole Lotta Scale: Faithful LLM Training Emulation with CrystalLLM
A Few GPUs, A Whole Lotta Scale: Faithful LLM Training Emulation with CrystalLLM.…
- 5 Semantic Scholar
TuxBot: Semantic-Aware Online OS Tuning with LLMs
TuxBot: Semantic-Aware Online OS Tuning with LLMs.…
- 6 Semantic Scholar
MCon: Building an Elastic Cloud for Android Applications with Framework Consolidation
MCon: Building an Elastic Cloud for Android Applications with Framework Consolidation.…
- 7 Semantic Scholar
MeshRT: Compile-Time Governed Wafer-Scale Runtime for Low-Latency High-Throughput Inference
MeshRT: Compile-Time Governed Wafer-Scale Runtime for Low-Latency High-Throughput Inference.…
- 8 Semantic Scholar
Janus: Multi-LLM Serving at Production Scale
Janus: Multi-LLM Serving at Production Scale.…
- 9 Semantic Scholar
I3DP: Neuro-Symbolic Inductive Invariant Inference for Distributed Protocols
I3DP: Neuro-Symbolic Inductive Invariant Inference for Distributed Protocols.…
- 10 Semantic Scholar
DiFlow: A System for Micro-Serving Text-to-image Diffusion Workflows
DiFlow: A System for Micro-Serving Text-to-image Diffusion Workflows.…
- 11 Semantic Scholar
Development and validation of a pragmatic pipeline for clinical free-text annotation using locally deployed open-weight large language models
Development and validation of a pragmatic pipeline for clinical free-text annotation using locally deployed open-weight large language models. Scientific Report…
- 12 Semantic Scholar
Anchor: Mitigating GPU Shallow Disruptions with Decoupled Memory
Anchor: Mitigating GPU Shallow Disruptions with Decoupled Memory.…
- 13 Semantic Scholar
Exploring how AI coaching promotes emotional intelligence awareness of contact centre team leaders
Team leaders in high-pressure contact centres must balance operational demands with people management, requiring high emotional intelligence (EQ). While traditi…
- 14 Semantic Scholar
Taming Dynamism on GPUs: Cross-SM Kernel Fusion via SM Cooperation and Just-in-Time Reduction
Taming Dynamism on GPUs: Cross-SM Kernel Fusion via SM Cooperation and Just-in-Time Reduction.…
- 15 Semantic Scholar
Validating Rust Compilers with Trait-Type Constraint Graph
Validating Rust Compilers with Trait-Type Constraint Graph.…
- 16 Semantic Scholar
PLS-ZO: projected latent subspace zeroth-order optimization for parameter-efficient fine-tuning
PLS-ZO: projected latent subspace zeroth-order optimization for parameter-efficient fine-tuning. International Journal of Machine Learning and Cybernetics…
- 17 Semantic Scholar
Guided-Grounded-InstructPix2Pix (GGIP2P): Precise targeting and spatially guided control for instruction-based image editing
Guided-Grounded-InstructPix2Pix (GGIP2P): Precise targeting and spatially guided control for instruction-based image editing. Multimedia tools and applications…
- 18 Semantic Scholar
Using large language models to analyze student reflections on generative AI in graduate coursework
Using large language models to analyze student reflections on generative AI in graduate coursework. Discover Education…
- 19 Semantic Scholar
Linux AGX: An Adaptive GPU eXtension to Linux Fair Scheduling for Physical AI and Robotic Systems
Linux AGX: An Adaptive GPU eXtension to Linux Fair Scheduling for Physical AI and Robotic Systems.…
- 20 Semantic Scholar
Understanding how expertise shapes human-AI collaboration in graphic design with PatternPursuit
Understanding how expertise shapes human-AI collaboration in graphic design with PatternPursuit. Design Science…
- 21 Semantic Scholar
Batched in Back: Characterizing and Optimizing Offline LLM Inference in Production with ACDC
Batched in Back: Characterizing and Optimizing Offline LLM Inference in Production with ACDC.…
- 22 Semantic Scholar
LMTracer: Fine-Grained and Real-Time Performance Profiling for Production LLM Systems
LMTracer: Fine-Grained and Real-Time Performance Profiling for Production LLM Systems.…
- 23 Semantic Scholar
Agentic AI-enabled influencers and psychological persuasion: agenticity shaping emotional engagement and well-being
The purpose of this study is to examine the psychological and societal effects of artificial intelligence-enabled influencers, focusing on how perceived authent…
- 24 Semantic Scholar
TensorDex: A Compact, Tensor-Centric Storage System for Modern AI Models
TensorDex: A Compact, Tensor-Centric Storage System for Modern AI Models.…
- 25 Semantic Scholar
Creative potential of image-generative AI models for conceptual engineering design tasks
Creative potential of image-generative AI models for conceptual engineering design tasks. Artificial intelligence for engineering design, analysis and manufactu…
- 26 Semantic Scholar
Skill VM: Write Once, Run Everywhere Efficiently
Skill VM: Write Once, Run Everywhere Efficiently.…
- 27 arXiv
FurE: Efficient Instance-Specific 3D Fur Reconstruction without Animal-Fur Datasets
Realistic and editable animal fur reconstruction from multi-view images is challenging due to fine-scale detail, self-occlusion and obfuscation, and, unlike hum…
- 28 arXiv
Telescopic Language Models
One deployed language model must often serve many compute budgets, yet serving each budget still means a separate training or compression run per point. We trai…
- 29 arXiv
PDMD: Projected Distribution Matching Distillation for Video Diffusion Models
Modern video diffusion models require tens of denoising evaluations over long spatiotemporal token sequences. Distribution Matching Distillation (DMD) reduces t…
- 30 arXiv
Retrieving Biblical Intertextual References in Karen Blixen's Seven Gothic Tales
Identifying intertextual references is central to literary scholarship, but computationally difficult when source material is transformed through paraphrase, al…
2026-09-28 · 週一 10 篇
- 1 👍 118
FuseReg: Regularizing Layer Fusion Mitigates the Reconstruction-Generation Gap in Representation Autoencoders
Representation autoencoders (RAEs) reuse features from a pretrained visual encoder as reconstruction and diffusion latents, integrating strong visual representa…
- 2 👍 64
Disaggregated Quantization: Specializing LLM Prefill and Decode
Prefill and decode reward different approaches to quantization: low-precision arithmetic accelerates prompt processing, while compact weights reduce memory traf…
- 3 👍 47
RayOrch: Programming and Executing Lineage-Controlled Multi-Grain Dataflows for Foundation-Model Data Preparation
Preparing high quality training data for foundation models requires scalable pipelines that transform heterogeneous documents and videos into structured records…
- 4 👍 22
Block Sparse Attention with Log-Linear Complexity
Scaling language models to long contexts is limited by the quadratic cost of self-attention. Block sparse attention offers an efficient alternative, but selecti…
- 5 👍 19
InternW0-Δ: A World Action Model Bridging Predictive Dynamics and Actions with 20K+ Hours of Open Data
World Action Models (WAMs) jointly model visual dynamics and action generation for generalist robot manipulation. A central challenge is to integrate priors fro…
- 6 👍 13
Tactile-JEPA: Topology-Aware Self-Supervised Representation Learning for Distributed Tactile Sensors
Tactile sensing is an essential modality for robots performing contact-rich, dexterous manipulation, particularly under visual occlusion. While pre-trained imag…
- 7 👍 10
FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance
Reference-based image quality assessment (IQA) metrics aim to reflect how humans perceive the perceptual distance between a pair of images. To learn how the hum…
- 8 👍 9
Enhancing Photogrammetric Digital Surface Models with Pretrained Diffusion Models and Multimodal Conditioning
Large-scale Digital Surface Models (DSMs) can be produced cost-effectively from satellite images via stereo-photogrammetry. However, the resulting 3D maps are o…
- 9 👍 9
Jev in the Wild: A Data-Driven Analysis of the Jev Model's Functionality, Applications and Ecosystem
Jev is a fast, low-cost decision model that answers natural-language questions with choices, binary judgments, and scores. As its public ecosystem grows rapidly…
- 10 👍 8
AgentWorld: Benchmarking Long-Horizon Collaboration of Multi-agent LLMs
Existing multi-agent benchmarks primarily test in competitive settings, short-horizon interactions under 20 steps, or simply aggregate individual performance, f…
更多論文（arXiv / ChatPaper / Semantic Scholar，26 篇）
- 1 Semantic Scholar
Towards Tractable Hardware Model Validation Against Real Hardware
Verifying low-level system code requires reasoning about how software interacts with the hardware environment on which it runs. Typically, this means specifying…
- 2 Semantic Scholar
Pwnagent: a knowledge-guided multi-agent system for automatic exploit generation
Pwnagent: a knowledge-guided multi-agent system for automatic exploit generation. Cybersecurity…
- 3 Semantic Scholar
The shifting landscape of assessment in STEM education in the age of generative AI
The shifting landscape of assessment in STEM education in the age of generative AI. International Journal of STEM Education…
- 4 Semantic Scholar
TierSizer: Counterfactual Reasoning for DRAM Sizing in Tiered Memory Systems
Compute Express Link (CXL) enables cost-effective memory capacity expansion by placing additional tiers behind a coherent fabric. To overcome the performance lo…
- 5 Semantic Scholar
LLM-driven embodied intelligent system for autonomous open-channel hydraulics: design and validation via long-duration PIV measurements
LLM-driven embodied intelligent system for autonomous open-channel hydraulics: design and validation via long-duration PIV measurements. Environmental Fluid Mec…
- 6 Semantic Scholar
A Survey of conversational AI from rule based to generative and retrieval augmented generation chatbots
A Survey of conversational AI from rule based to generative and retrieval augmented generation chatbots. Discover Artificial Intelligence…
- 7 Semantic Scholar
Do “AI-assisted” labels reduce perceived quality of scenarios? An exploratory experimental study
Do “AI-assisted” labels reduce perceived quality of scenarios? An exploratory experimental study. European Journal of Futures Research…
- 8 Semantic Scholar
Towards music-aware robot dance generation: bridging rhythm, motion and control
Towards music-aware robot dance generation: bridging rhythm, motion and control. Multimedia Systems…
- 9 Semantic Scholar
A style aware attention transformer for fine grained interior design style recognition
A style aware attention transformer for fine grained interior design style recognition. Discover Artificial Intelligence…
- 10 Semantic Scholar
Empirical Analysis of Goal Hijacking in Large Vision-Language Models via Visual Prompt Injection
Empirical Analysis of Goal Hijacking in Large Vision-Language Models via Visual Prompt Injection. New generation computing…
- 11 Semantic Scholar
Dynamic risk budgeting via large language models: A context-aware framework for adaptive portfolio allocation
Dynamic risk budgeting via large language models: A context-aware framework for adaptive portfolio allocation. Applied intelligence (Boston)…
- 12 Semantic Scholar
Fine-tuning large language models for binary and multiclass target-specific hate speech detection
Fine-tuning large language models for binary and multiclass target-specific hate speech detection. Discover Artificial Intelligence…
- 13 Semantic Scholar
From epistemic promises to real world practices: the ongoing becoming of 3D models and urban digital twins in municipal planning
From epistemic promises to real world practices: the ongoing becoming of 3D models and urban digital twins in municipal planning. Computational Urban Science…
- 14 Semantic Scholar
SceneContextNet: an integrated multi-stage deep learning framework for context-aware scene understanding using multi-object detection and spatial relationship modelling
SceneContextNet: an integrated multi-stage deep learning framework for context-aware scene understanding using multi-object detection and spatial relationship m…
- 15 Semantic Scholar
Editorial
Dear Readers, It gives me great pleasure to announce the tenth issue of 2026. In this issue, 6 articles by 16 authors from 6 countries – Belgium, China, Czech…
- 16 Semantic Scholar
Automating Clavien–Dindo classification with large language models in percutaneous nephrolithotomy: an exploratory study
Automating Clavien–Dindo classification with large language models in percutaneous nephrolithotomy: an exploratory study. World journal of urology…
- 17 Semantic Scholar
Hydra: Rich and Scalable Functional Verification of eBPF Deployments
Production eBPF deployments now run as deep, dynamically composed chains of programs. Reasoning about the behavior of such chains is critical - e.g., an upstrea…
- 18 Semantic Scholar
Visual narratives of disaster on Instagram: a multimodal AI-assisted analysis
Visual narratives of disaster on Instagram: a multimodal AI-assisted analysis. Computational Urban Science…
- 19 Semantic Scholar
Simulated Morality, Misplaced Trust: The Risks of Treating AI as a Moral Partner
Simulated Morality, Misplaced Trust: The Risks of Treating AI as a Moral Partner. Philosophy & Technology…
- 20 Semantic Scholar
AI-assisted retrospective reconstruction of contested psychiatric care: an autoethnographic case study
AI-assisted retrospective reconstruction of contested psychiatric care: an autoethnographic case study. AI and Ethics…
- 21 Semantic Scholar
Unlocking the value of generative AI for supply chain resilience and crisis readiness: a complex adaptive systems perspective
The existing literature conceptualizes generative artificial intelligence (GenAI) as a knowledge infrastructure that facilitates organizational abilities to use…
- 22 Semantic Scholar
CycleSDE-AD: a score-based SDE framework for missing-robust anomaly detection in industrial air compressor time series
CycleSDE-AD: a score-based SDE framework for missing-robust anomaly detection in industrial air compressor time series. Journal of Supercomputing…
- 23 Semantic Scholar
SynthCoder: Anti-pattern identification and model training for FIM mode code completion
SynthCoder: Anti-pattern identification and model training for FIM mode code completion. Empirical Software Engineering…
- 24 Semantic Scholar
Human-Annotated or MLLM-Assisted? A Comparative Evaluation of Filmic Metaphor Identification Using FILMIP-AI
Abstract This study introduces FILMIP-AI, an operationalization of the Filmic Metaphor Identification Procedure (FILMIP) for multimodal large language models (M…
- 25 Semantic Scholar
Digital technology risk exposure and corporate key core technology innovation
Digital technology risk exposure and corporate key core technology innovation. Digital Economy and Sustainable Development…
- 26 Semantic Scholar
CrossServe: Cross-Layer Scheduling for SLO Optimization in Multi-Tenant LLM Serving
The deployment of Large Language Models (LLMs) as multi-tenant cloud services is now widespread, but maintaining high Service Level Objective (SLO) attainment a…
2026-09-27 · 週日 0 篇
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
MuRS 2026: 4th Music Recommender Systems Workshop
Music recommendation has been a core area of interest within the recommender systems community since its early days. With the rise of music streaming platforms,…
- 2 Semantic Scholar
Efficient Ambiguity Resolution via Information-Guided Clarification for Text-to-SQL
Text-to-SQL technology enables users to query relational databases using natural language, but ambiguity in user questions remains a major obstacle: non-expert …
- 3 Semantic Scholar
From Side Information and Knowledge Graphs to Large Language Models: Two Decades of Knowledge Integration in Recommender Systems
Over the past two decades, Recommender Systems (RSs) have repeatedly changed how they integrate external knowledge, moving from constraints and side information…
- 4 Semantic Scholar
Does AI reduce the reputation premium in knowledge diffusion?
Does AI reduce the reputation premium in knowledge diffusion?. Marketing letters…
- 5 Semantic Scholar
On the Way to Explainable Course Recommendation: Enriching Graph-Based Modeling with Sequential Signals
Course recommendation systems (CRS) play a crucial role in guiding university students through their academic journey by assisting in course selection. Oftentim…
- 6 Semantic Scholar
Backward and Inward: From Preferences to Personalization and the Safety Crisis Within Conversational Recommender Systems
Conversational recommender systems (CRS) pursue ever-deeper personalization, dissolving the separation between system and user that earlier paradigms maintained…
- 7 Semantic Scholar
PLAIN: An Explainable Generative Search System Enhanced by Multi-granularity Semantic Alignment
Industrial search platforms must efficiently retrieve relevant items from billions of candidates while satisfying both query relevance and user preferences. Gen…
- 8 Semantic Scholar
Bridge the Unseen Gap: Enhancing Non-overlapping Cross-domain CTR Prediction via Profile Retrieval
Click-through rate (CTR) prediction is a fundamental task in industrial recommender systems. Cross-domain CTR prediction, which leverages data from a source dom…
- 9 Semantic Scholar
How Faithful Is the Reasoning of LLM Recommenders? A Counterfactual Audit
Large language model (LLM)-based recommenders increasingly generate reasoning alongside their recommendations [5], implicitly suggesting that the stated reasoni…
- 10 Semantic Scholar
Safety-Aware Next-POI Recommendation with Large Language Models
Point of Interest (POI) recommendation has become a core task in location-based services, with modern systems increasingly driven by deep learning models that a…
- 11 Semantic Scholar
RPSC: Robust LLM Scheduling by Tolerating Prediction Inaccuracy and Mitigating Tail Latency
High-concurrency LLM serving requires inference systems to balance execution efficiency with stringent performance guarantees. While Shortest Remaining Processi…
- 12 Semantic Scholar
In-Batch Negatives Can Silently Cripple LLM-Encoded Sequential Recommenders
LLM-based sequential recommenders increasingly train user/item encoders with in-batch InfoNCE negatives inherited from contrastive learning—convenient, but it s…
- 13 Semantic Scholar
PROMISE: Process Reward Models for Unlocking Test-Time Scaling Laws in Generative Recommendations
Generative Recommendation has emerged as a promising paradigm, reformulating recommendation as a sequence-to-sequence generation task over hierarchical Semantic…
- 14 Semantic Scholar
Intelligent diagnosis method of English learning disabilities based on knowledge mapand long-term and short-term memory network
Intelligent diagnosis method of English learning disabilities based on knowledge mapand long-term and short-term memory network. Discover Artificial Intelligenc…
- 15 Semantic Scholar
Edge-Cloud Collaborative RAG with Multi-State Semantic Caching
Enterprise applications are increasingly adopting collaborative Retrieval Augmented Generation (RAG) frameworks that synergize Small Language Models (SLMs) at t…
- 16 Semantic Scholar
User-Controlled Intent Layers for LLM-Mediated Personalization: A Research Agenda for Recommender Systems
Recommender systems deployed at scale are predominantly organized around platform-centric architectures in which each service independently constructs and optim…
- 17 Semantic Scholar
Online Scheduling of Battery-Aware Speculative Decoding for Energy-Efficient Cloud-Edge Collaborative LLM Inference
While distributed speculative decoding can offer efficient acceleration for Large Language Model (LLM) inference in cloud-edge environments, unleashing its full…
- 18 Semantic Scholar
FIBB: Capturing False Information Beliefs and Behaviors Across Physical and Online Environments
Humans assess the veracity of newly presented false information using multiple sources: from observations in the online environment (e.g., social media content …
- 19 Semantic Scholar
MIGServe: Layout-Aware Multi-Instance GPU Management for Efficient LLM Serving
NVIDIA’s Multi-Instance GPU (MIG) technology partitions a single GPU into hardware-isolated instances of varying sizes, offering a promising substrate for servi…
- 20 Semantic Scholar
Unified Search and Recommendation Workshop
Search (IR) & recommendation (RecSys) have long been developed in parallel, despite serving the same discovery goal over shared item universes. The rise of LLMs…
- 21 Semantic Scholar
What Reasoning Should a Recommender Have? Designing Reasoning for Inference-Time Scaling in Recommendation
Large language models can gain accuracy by using more computation at inference, most visibly through chain-of-thought, where the model proceeds step by step in …
- 22 Semantic Scholar
DPGFlow: Decoupled Preference Guided Flow Matching for Cross Domain Sequential Recommendation
Cross-Domain Sequential Recommendation (CDSR) aims to improve next-item prediction by leveraging users’ sequential behaviors across multiple domains. Despite re…
- 23 Semantic Scholar
Matching granularity matters for occupation skill and qualification linking in job advertisements
Matching granularity matters for occupation skill and qualification linking in job advertisements. Discover Artificial Intelligence…
- 24 Semantic Scholar
The Third Workshop on Agentic and Generative AI for E-Commerce
The Third Workshop on Agentic and Generative AI for E-Commerce, co-located with the 20th ACM Conference on Recommender Systems (RecSys 2026), brings together re…
- 25 Semantic Scholar
Codebook-Based Semantic IDs in Generative Recommendation: Enabling Interface, Emerging Bottleneck
Codebook-based semantic IDs (SIDs), short discrete code sequences produced by quantization over item representations, made generative recommendation practical b…
- 26 Semantic Scholar
CEDAR: Compressed-Object Parallel Retrieval for Scalable AI Systems
Retrieval-augmented AI systems maintain external memory through parallel pipelines that shard corpora, batch embeddings, build ANN indexes, refresh updates, and…
- 27 Semantic Scholar
Forward and Outward: From Aggregate to Individual and Towards Human Understanding in Conversational Recommendation
Conversational recommender systems (CRS) have made substantial progress leveraging LLMs for preference elicitation and recommendation. Yet recent evidence sugge…
- 28 Semantic Scholar
Helios: Exposing Context Lifecycles for Memory-Efficient Agentic LLM Serving
LLM serving is shifting from stateless queries to long-running agents that reshape working context through consolidation, updates, and forgetting. Existing serv…
- 29 Semantic Scholar
RecTemp: Temporal Reasoning in Recommendation Systems
User preferences, intents, behaviors, and contexts evolve over time, making temporal reasoning a fundamental challenge in recommender systems. Accurately captur…
- 30 Semantic Scholar
WAQ-LLM: Optimizing Multi-Instance LLM Deployment via Workload-Aware Queueing Model
The deployment of Large Language Models (LLMs) on multi-instance GPU clusters has become essential to meet the surging demand for generative AI applications. Wh…
2026-09-26 · 週六 0 篇
更多論文（arXiv / ChatPaper / Semantic Scholar，16 篇）
- 1 Semantic Scholar
From flip to click: a global bibliometric analysis of blended learning trends in programming education (2000–2023)
From flip to click: a global bibliometric analysis of blended learning trends in programming education (2000–2023). SN Social Sciences…
- 2 Semantic Scholar
Trusted and explainable collective learning for conflictive multi-view decision-making
Trusted and explainable collective learning for conflictive multi-view decision-making. Knowledge and Information Systems…
- 3 Semantic Scholar
Abduction In Silico: Large Language Models Replicate Human Inference to the Best Explanation
Abduction In Silico: Large Language Models Replicate Human Inference to the Best Explanation. Minds and Machines…
- 4 Semantic Scholar
An integrative theoretical model of AI-supported cybersecurity governance and organizational resilience
An integrative theoretical model of AI-supported cybersecurity governance and organizational resilience. Discover Artificial Intelligence…
- 5 Semantic Scholar
Improving patient communication and understanding of medical reports through text simplification with ChatGPT: a comparative study
Improving patient communication and understanding of medical reports through text simplification with ChatGPT: a comparative study. Oral and Maxillofacial Surge…
- 6 Semantic Scholar
Dance arrangement and style transfer methods driven by diffusion models and spatiotemporal graph convolutional networks
Dance arrangement and style transfer methods driven by diffusion models and spatiotemporal graph convolutional networks. Discover Artificial Intelligence…
- 7 Semantic Scholar
AI-driven organisational cognitive transformation: knowledge management strategies for hybrid cognition
This paper aims to position artificial intelligence (AI) as a knowledge-based cognitive technology that drives organisational cognitive transformation (OCT). Th…
- 8 Semantic Scholar
Extending the technology acceptance model with awareness and self-efficacy in predicting ChatGPT adoption among Nigerian marketing students
Extending the technology acceptance model with awareness and self-efficacy in predicting ChatGPT adoption among Nigerian marketing students. Discover Education…
- 9 Semantic Scholar
Business clusters and SMEs' sustainable performance in emerging markets: examining the roles of open innovation and digital capacity
This study aims to develop a framework for the sustainable performance (SP) of small and medium enterprises (SMEs) that will benefit emerging markets by emphasi…
- 10 Semantic Scholar
LENS-LLM Effects on News Scrutiny: Influence, Delegation, and Memory in AI-Mediated News.
LENS-LLM Effects on News Scrutiny: Influence, Delegation, and Memory in AI-Mediated News.. Cyberpsychology, Behavior, and Social Networking…
- 11 Semantic Scholar
Generative AI Utilization and Perspectives Among Reproductive-Age Women with Systemic Autoimmune and Rheumatic Diseases.
Generative AI Utilization and Perspectives Among Reproductive-Age Women with Systemic Autoimmune and Rheumatic Diseases.. Arthritis care & research…
- 12 Semantic Scholar
Automated formalization and verification of building code requirements using knowledge graph construction
This study develops and evaluates a methodology for automated formalization and verification of building code requirements using ontology-based knowledge graphs…
- 13 Semantic Scholar
Automated grading of short-answer image-based assessments using a hybrid natural language processing-AI framework: validation in radiology education.
Automated grading of short-answer image-based assessments using a hybrid natural language processing-AI framework: validation in radiology education.. Singapore…
- 14 Semantic Scholar
How generative AI shapes the last meal at life's end.
How generative AI shapes the last meal at life's end.. Nursing Ethics…
- 15 Semantic Scholar
Artificial intelligence use and its influence on academic learning behaviour among undergraduate students in Nigeria
Artificial intelligence use and its influence on academic learning behaviour among undergraduate students in Nigeria. Discover Education…
- 16 Semantic Scholar
Impact of AI assistance on pharmacist and pharmacy trainee accuracy and automation bias during medication order verification: A randomized simulation trial.
Impact of AI assistance on pharmacist and pharmacy trainee accuracy and automation bias during medication order verification: A randomized simulation trial.. Am…
2026-09-25 · 週五 10 篇
- 1 👍 195
Training Object Permanence in World Models
Object permanence and solidity are hallmarks of human cognitive priors. Recent studies show that video generation models, a paradigmatic class of current world …
- 2 👍 64
Your Transformer Can Hold Two Thoughts at Once: Evidence of Linear Superposition in LLMs
While Large Language Models (LLMs) rely on highly non-linear components, in this work we demonstrate that they exhibit fundamental linearity: when inputs from d…
- 3 👍 32
WanPE: Towards Cinematic Prompt Enhancement for Modern Text-to-Video Generation
Video generation begins in text space by authoring a cinematic screenplay, then materializes into pixels. As contemporary video generators scale to 30 seconds a…
- 4 👍 20
OmniEcho: Spatial Audio Understanding for Embodied Agents
Humans can effortlessly localize the direction of a sound source and integrate it with visual cues for reasoning, yet this remains challenging for embodied agen…
- 5 👍 12
Rufus-Air: An Open LLM Post-Training Recipe
Rufus-Air is an open and reproducible post-training recipe on GLM-4.5-Air-Base (106B-A12B), organized as a serial pipeline of eight stages: SFT, Reasoning RL, C…
- 6 👍 10
Parts-of-Speech as Emergent Categories in SAE Latent Space
Sparse AutoEncoders (SAEs) offer a promising way to inspect language model representations, but it is still unclear what kind of linguistic structure their late…
- 7 👍 10
Qwen-Planner-Agent: A Closed-Loop AI-for-AI Framework for Real-World Mobile Planner Agents
The rapid progression of large language models is extending AI from passive content generation into the active workflows of engineering and scientific discovery…
- 8 👍 9
IterSynth: Rethinking Deep Search Agents via Role-Decoupled Iterative Synthesis
Deep search requires LLM agents to decompose complex queries, search for evidence, and synthesize grounded answers, yet existing ReAct-style agents suffer from …
- 9 👍 9
Coding Agents for Generalized Task and Motion Planning Problems
Task and motion planning (TAMP) problems remain difficult even with full observability and object-centric states because discrete decisions are tightly coupled …
- 10 👍 9
ExplorationBench: Measuring AI Systems' Exploration in Verifiable Alien Worlds
Scientific discovery begins where known problems end. There, AI systems must engage in exploration: framing hypotheses, designing experiments, and iterating on …
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Artificial intelligence in science education with a focus on life and earth sciences a bibliometric analysis of global research trends from 2015 to 2025
Artificial intelligence in science education with a focus on life and earth sciences a bibliometric analysis of global research trends from 2015 to 2025. Discov…
- 2 Semantic Scholar
Bit-sliced quantization for robust memristor-based compute-in-memory
Bit-sliced quantization for robust memristor-based compute-in-memory. npj Unconventional Computing…
- 3 Semantic Scholar
New Developments in Dental Education, Research, and Clinical Practice Opened by Generative AI - Evaluation of Large Language
New Developments in Dental Education, Research, and Clinical Practice Opened by Generative AI - Evaluation of Large Language. The Journal of the Kyushu Dental S…
- 4 Semantic Scholar
Developing research reasoning through digital tools in PhD students’ methodological training
Developing research reasoning through digital tools in PhD students’ methodological training. Discover Education…
- 5 Semantic Scholar
Are generative AI-facial composites of suspects too good? Accurate facial composites decrease memory strength for suspects
This study aims to examine the artificial intelligence (AI)-witness effect, a recently observed phenomenon demonstrating impaired recognition performance for su…
- 6 Semantic Scholar
Artificial intelligence, religion, and society: ethical leadership and the shaping of value-based adaptive governance
Artificial intelligence, religion, and society: ethical leadership and the shaping of value-based adaptive governance. AI and Ethics…
- 7 Semantic Scholar
MFLab: decoupled luminance-chrominance modeling with uncertainty-aware flow refinement for low-light image enhancement
MFLab: decoupled luminance-chrominance modeling with uncertainty-aware flow refinement for low-light image enhancement. Journal of King Saud University: Compute…
- 8 Semantic Scholar
Integrating Large Language Models (LLMs) with Oracle 26AI for Advanced Enterprise Analytics and Knowledge Management
Enterprise organizations increasingly hold two categories of information assets that have historically been managed by incompatible systems: structured relation…
- 9 Semantic Scholar
A systematic review of attention duality in deep visual models for performance and faithfulness tradeoffs
A systematic review of attention duality in deep visual models for performance and faithfulness tradeoffs. Discover Artificial Intelligence…
- 10 Semantic Scholar
From card catalogs to chatbots: envisioning academic library services in the era of AI and LLMs
This article introduces the CARDLG framework: Collections, Access, Retrieval, Description, Literacy and Governance. The framework analyzes how generative artifi…
- 11 Semantic Scholar
Mal-Intelligence: how artificial intelligence produces a sustained illusion of cognitive competence
Mal-Intelligence: how artificial intelligence produces a sustained illusion of cognitive competence. AI and Ethics…
- 12 Semantic Scholar
From capability to assertability: epistemic regulation of LLM-mediated claim presentation
From capability to assertability: epistemic regulation of LLM-mediated claim presentation. Ethics and Information Technology…
- 13 Semantic Scholar
Conflict-aware commonsense knowledge distillation with hypergraph reasoning for robust multimodal sentiment analysis
Conflict-aware commonsense knowledge distillation with hypergraph reasoning for robust multimodal sentiment analysis. Journal of King Saud University: Computer …
- 14 Semantic Scholar
AI-based smart load balancing for sustainable electric vehicle charging hubs using LSTM forecasting and reinforcement learning
AI-based smart load balancing for sustainable electric vehicle charging hubs using LSTM forecasting and reinforcement learning. Discover Electronics…
- 15 Semantic Scholar
Analyzing and fixing code generation errors of foundation large language models
Analyzing and fixing code generation errors of foundation large language models. Empirical Software Engineering…
- 16 Semantic Scholar
SATIN: a sparse attention application-specific instruction-set processor with custom mixed-precision extensions for energy-efficient edge inference of transformer large language models
SATIN: a sparse attention application-specific instruction-set processor with custom mixed-precision extensions for energy-efficient edge inference of transform…
- 17 arXiv
LLM Agents Can Easily Tamper With Their Own Traces
Asynchronous monitoring, incident investigations, and compliance audits primarily rely on agent traces to reconstruct what happened. These analyses assume that …
- 18 arXiv
AD-WM: Action-Discriminative World Models for Counterfactual Model Predictive Control
Latent world models are typically trained to predict factual transitions, whereas model predictive control (MPC) must compare alternative actions from the same …
- 19 arXiv
Temporal Gradient Inversion for Private Trajectory Reconstruction in Embodied Reinforcement Learning
Distributed learning in embodied reinforcement-learning agents offers a degree of privacy by retaining raw sensor data on-device and transmitting only policy gr…
- 20 arXiv
Agentic Detection of Online Conspiracies
Conspiratorial discourse on social media is not always expressed through explicit claims or stable lexical markers. The same surface content may express endorse…
- 21 arXiv
RAPID: Robot Agentic Programming from Demonstrations
Coding agents have demonstrated enormous success in solving complex programming problems. To leverage their potential for robot systems, this work introduces Ro…
- 22 arXiv
Rolling-WAM: World Action Models with Rolling Imagination
World Action Models (WAMs) couple action generation with future visual prediction for robotic manipulation. However, completing the joint video-action denoising…
- 23 arXiv
Towards Practical Compression of 3D Gaussian Splatting
3D Gaussian Splatting (3DGS) enables high-quality novel-view synthesis but requires substantial storage. Existing compression methods often rely on spatial cont…
- 24 arXiv
JevOut: Natural Context Can Flip Decision Models
Dedicated decision models such as Jev map unstructured language to probability distributions over finite choices, allowing their outputs to directly route reque…
- 25 arXiv
SemMSA: Latent Semantic-Aided Robust Multimodal Sentiment Analysis with Incomplete Data
Recent research on Multimodal Sentiment Analysis (MSA) has focused on learning from language, visual, and acoustic modalities with incomplete data to infer huma…
- 26 arXiv
OmniFabric: Coherent UV Space Texture Synthesis for 3D Garment Reconstruction
Automated generation of production-ready 3D garment assets from a single image is a central challenge in digital content creation. While recent generative model…
- 27 arXiv
To Trust or Not to Trust: Retrieval-Augmented Fact Checking in Speech
Online misinformation increasingly appears in spoken formats such as news clips, podcasts, interviews, political speeches, and social media videos, creating a n…
- 28 arXiv
PoEM: Predicting RL Outcomes from Existing Policies
Foundation models are post-trained with reinforcement learning (RL) to maximize specific rewards, such as human alignment, correctness, or instruction following…
- 29 arXiv
BiCC: Bidirectional Connected-Component Loss for Instance-Aware Segmentation
Common segmentation losses aggregate errors voxel-wise, so lesions influence the objective in proportion to their volume, giving small but clinically critical l…
- 30 arXiv
TrackEverything: Long Horizon Dense Tracking via De-Duplicating 3D Scene Representations
Existing point tracking models face a fundamental tradeoff: they can either track a sparse set of query points over long horizons, or track all points across on…
2026-09-24 · 週四 10 篇
- 1 👍 43
Spatial-Interactor: Learning Spatial Reasoning through Interaction with the Observable Physical World
Spatial reasoning is essential for vision-language models (VLMs) to understand and act in the physical world. Reasoning in dynamic environments requires VLMs to…
- 2 👍 40
HappyWorld-Bench
Evaluating world models requires assessing both the quality of the worlds they generate and their consistency and responsiveness under exploration, interaction,…
- 3 👍 36
The Past Frames the Future: Memory for Autoregressive Video Generation
Advances in generative models have improved video fidelity, enabling long-horizon generation, interactive world modeling, and evolving visual environments. Auto…
- 4 👍 34
Just-in-Time Memory: Learning to Curate Task-Adaptive Memory for LLM Agents
Agentic memory systems reuse past experience to improve future performance, yet most existing designs curate memory at write time: once a task is completed, its…
- 5 👍 22
RewardVerse: Rubric-Guided Policy Optimization for Video Reward Modeling
Reinforcement learning (RL) is vital for optimizing video generation models, with a robust reward model (RM) serving as the cornerstone. However, existing video…
- 6 👍 16
Schrödinger's Code Repository: Have LLMs Learned SWE-bench or Memorized It?
Repository-level coding benchmarks have become the standard for evaluating coding agents, yet they inherently suffer from data leakage because they are built up…
- 7 👍 15
PACT: From Credit Assignment to Critic Alignment
Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted mathematica…
- 8 👍 13
GeoPair: Geometry-Preserving Cross-Layer Factorization for Training-Free Transformer Compression
Transformer architectures exhibit cross-layer redundancies, yet post-training compression pipelines typically optimize layers in isolation or rely on heuristic …
- 9 👍 10
PackLab: A Comprehensive Framework for Developing, Training, and Evaluating MLLMs in Robotic Bin Packing
Robotic bin packing requires long-horizon sequential decision-making, as each object placement affects the available space for subsequent packing. Existing meth…
- 10 👍 9
MemBodied: Recurrent Associative Memory for Vision-Language-Action Models
Vision-Language-Action models provide a strong foundation for general-purpose robot control, yet a vast majority of policies do not preserve and leverage episod…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 arXiv
Contrastive Learning for Authorship Verification
Our results show that contrastive learning outperforms a classification-based approach to authorship verification under the tested settings. We identify loss fu…
- 2 arXiv
Where Should I Join? Robot Group Joining via Language-Guided Goal Prediction
Social navigation typically assumes a specified goal and focuses on reaching it while respecting social conventions, whereas robot group joining requires predic…
- 3 arXiv
Even Sharper Bounds for Transductive Learning and Its Applications
We introduce Sharper Transductive Local Complexity (STLC), a localized complexity method for transductive learning under uniform sampling without replacement. T…
- 4 arXiv
Can LLMs Reason About Runtime Behavior? A Repository-Level Dynamic Benchmark
Large language models (LLMs) are increasingly used in coding tasks, but their ability to reason about code execution remains unclear. Existing repository-level …
- 5 arXiv
Nonequilibrium Phases of Repulsive Self-Attention: Chaos, Attention Condensation, and Emergent Locality
We study the nonequilibrium dynamics of a minimal recurrent transformer with $N$ normalized tokens, $Q=K=I$, and a negative value map $V=-I$. Similarity-based a…
- 6 arXiv
Order-Invariant Answers, Order-Sensitive Representations in Mathematical Reasoning
Reordering a set of mathematical rules without changing its meaning should preserve the correct answer, but must a model's internal representations stay invaria…
- 7 arXiv
HaRP: High Dynamic Range Photosequencing through Dual Reversed Shutter Scanning
The adoption of CMOS sensors in mobile photography is frequently compromised by the rolling shutter (RS) effect, which introduces geometric distortions and moti…
- 8 arXiv
Minimal-Norm Univariate Two-Layer ReLU Classification: Exact Solutions and Global Optimality with Skip Connections
We study minimal-norm interpolation and $\ell_2$-regularized logistic-loss minimization for binary classification by univariate two-layer ReLU networks. We give…
- 9 arXiv
MultiVENT-Raw: A Benchmark for Retrieval and Reasoning over Raw Videos
Online information is increasingly consumed in video format. Much of this comes in the form of *raw video*: continuous footage taken on a cell phone, with a han…
- 10 arXiv
Predicting the Progression of Adolescent Idiopathic Scoliosis
Adolescent Idiopathic Scoliosis is defined as a lateral curvature of the spine that develops during adolescence, without known cause. The condition can result i…
- 11 arXiv
Cross-Scale Transfer Learning for Depression Severity Prediction: From PHQ-8 to HAMD-17 Across Languages and Clinical Paradigms
This work addresses continuous depression-severity score prediction from clinical interview transcripts under data scarcity. We propose a sequential low-rank ad…
- 12 arXiv
Context-Continuous Preference Learning for Exoskeleton Personalization
Personalizing exoskeleton assistance across operating conditions is constrained by the time and physical effort required to collect user feedback. We examined w…
- 13 arXiv
Repairability of Inexact Solvers in Recursive State Estimation with Machine Learning
Recursive state estimation often executes approximate numerical solutions inside a feedback loop, where highly accurate local steps do not guarantee better over…
- 14 arXiv
The Skin-Restricted Reinhard Transform:Uniqueness under a Lightness-Preserving Constraint
Catalog skin recolouring has to change pigment and leave shading alone. The classical Reinhard map does not make that split: it rescales lightness by the ratio …
- 15 arXiv
Agent-Editing World Model: Rethinking World Modeling for LLM Agents
Recent advances in large language models (LLMs) have enabled agents to tackle long-horizon tasks across diverse environments. To further improve agent performan…
- 16 arXiv
Frozen Flows Forget: Diagnosing and Restoring Lost Motion in a Latent-flow World Model
Latent world models that integrate a flow in a frozen self supervised latent space train stably and cheaply, yet silently lose the property manipulation depends…
- 17 arXiv
Learning Holographic Reduced Representations with Clifford Variational Autoencoders
Vector Symbolic Algebras project data structures into a hyperdimensional vector space through the application of their vector algebras to randomly generated ato…
- 18 arXiv
Learning Collective Dynamics with Differentiable Gaussian Representations
Collective responses depend on individual differences, contact opportunities, and accumulated experience. Learning their dynamics from aggregate counts requires…
- 19 arXiv
Memory Attention
Language models typically construct attention values from contextual hidden states, even when some of their content may be reusable across contexts. We investig…
- 20 arXiv
Fine-Tuning LLMs for Translation: General Forgetting Mitigation Does Not Preserve MT-Specific Instruction Following
Fine-tuning large language models on parallel data improves translation quality but can cause catastrophic forgetting. Mitigation methods are generally evaluate…
- 21 arXiv
Quantum score matching with applications to learning thermal states
Score matching has driven major advances in classical generative learning by enabling models to learn from data without evaluating intractable normalization con…
- 22 arXiv
When and Where to Trust the Teacher: Unifying On-Policy Distillation and GRPO through Entropy-Calibrated Credit Assignment
Reinforcement learning with verifiable rewards (RLVR) supervises mathematical reasoning through final-answer correctness, but provides little guidance on indivi…
- 23 arXiv
ForgetMimic: Motion Unlearning for Reinforcement Learning Humanoid Control
Humanoid control, leveraging human demonstrations, has achieved diverse, agile, and natural locomotion behaviors through reinforcement learning (RL). While this…
- 24 arXiv
Shopping by algorithm: How agentic AI deploys human heuristics as a surrogate consumer
Consumers increasingly delegate purchasing decisions to Large Language Models (LLMs) acting as surrogate consumers. Using "Tool-Lab," an adaptation of informati…
- 25 arXiv
AnchorReasoning: A Visual Grounding and Causal Reasoning Dataset in Long-Tail Autonomous Driving Scenarios
Vision-language models (VLMs) offer a promising approach to long-tail autonomous driving, but existing driving datasets provide limited supervision for connecti…
- 26 arXiv
LEAP-CBF: A Safety Filter for Uncertain Systems with Least-Effort Adversarial Potentials
Control barrier functions (CBF) are a popular safety filter to ensure safety for nonlinear dynamical systems. However, when the system is subject to uncertainti…
- 27 arXiv
Privacy-Preserving Semantic Segmentation from High-Resolution Depth and Ultra-Low-Resolution RGB
As mobile robots become increasingly integrated into everyday environments, privacy risks arising from onboard cameras have become a growing concern. Ultra-low-…
- 28 arXiv
MicroQonv: Reshaping Convolution Tensors for Efficient Microscaling in Training and Inference
Microscaling quantization techniques are increasingly used to represent neural network parameters with 8 bits or fewer while preserving near-full precision accu…
- 29 arXiv
Digital diglossia: Arabic between X and Facebook
This study highlights the distribution of Standard Arabic (SA; H(igh) variety) and Colloquial Arabic (CA; L(ow) variety) across X and Facebook. 16754 public pos…
- 30 arXiv
Mizar: A 159M-Parameter Audio-Language Model for Audio Understanding
Audio-language models (ALMs) integrate acoustic perception with the knowledge encoded in language models, enabling contextual understanding of auditory events. …
2026-09-23 · 週三 10 篇
- 1 👍 117
The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks
LLM agents increasingly work on long-horizon tasks, and the decisions they make along the way, such as which hypothesis to test or which implementation to build…
- 2 👍 65
RULER: Instance-aware Rubric Rewards for SVG Generation
Generating Scalable Vector Graphics (SVG) code from natural-language instructions is an open-ended task without absolute visual ground truth, leaving both evalu…
- 3 👍 39
All-in-One Multilingual Scene Text Recognition with Script-aware Mixture-of-Experts
Multilingual scene text recognition (STR) remains challenging due to the scarcity of training data for most languages and the difficulty of serving diverse scri…
- 4 👍 36
Ovis-Embedding: Pushing the Frontiers of Universal Omni-Modal Embeddings
In this report, we introduce Ovis-Embedding, a state-of-the-art omni-modal embedding family built on native integration of text, image, video, and audio. Instea…
- 5 👍 27
Bellman Policy Optimization
Reinforcement learning with verifiable rewards (RLVR) improves the reasoning capabilities of large language models (LLMs). We introduce Bellman Policy Optimizat…
- 6 👍 26
Circuit Hypernetworks for Quantum-Augmented Diffusion Language Models
Language models can be adapted by changing the computations applied to individual tokens. Quantum circuits offer one such approach, but evaluating wider circuit…
- 7 👍 25
StableVQ: Practical Guidelines for Stable Vector-Quantized Tokenizer Training
Vector Quantization (VQ) is fundamental to discrete visual tokenizers that power modern autoregressive and masked image generation models. While recent shared-p…
- 8 👍 25
JEV-as-a-Judge: Accept When Confident, Escalate When Unsure
LLM-as-a-judge enables evaluation across diverse tasks, but inference cost and confidence reliability become critical at scale. We study whether a decision-only…
- 9 👍 16
From Pattern Recognizers to Personalized Companions: A Survey of Large Language Models in Mental Health
The rising global prevalence of mental health conditions, together with longstanding barriers in traditional healthcare, such as limited resources, high cost, s…
- 10 👍 14
Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs
Diffusion Large Language Models (dLLMs) have recently emerged as a promising alternative to autoregressive LLMs by enabling non-autoregressive text generation. …
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar 📖 1
BD-Diff: Generative Diffusion Model for Image Deblurring on Unknown Domains with Blur-Decoupled Learning
BD-Diff: Generative Diffusion Model for Image Deblurring on Unknown Domains with Blur-Decoupled Learning. International Journal of Computer Vision…
- 2 Semantic Scholar
From Infrastructure Stewardship to Institutional Capability
In 2010, the University of Florida had eighty fragmented data centers—most of them closets—and computing infrastructure that could not support its ambition to b…
- 3 Semantic Scholar
From disclosure to verification: a risk-differentiated framework for the ethical use of AI in academic research
From disclosure to verification: a risk-differentiated framework for the ethical use of AI in academic research. Ethics and Information Technology…
- 4 Semantic Scholar
Beyond classification: a systematic review of advanced predictive methodologies in retinal image analysis
Beyond classification: a systematic review of advanced predictive methodologies in retinal image analysis. Research on Biomedical Engineering…
- 5 Semantic Scholar
Exploring young adult cancer survivors’ perspectives on generative AI chatbots for symptom support
Exploring young adult cancer survivors’ perspectives on generative AI chatbots for symptom support. Supportive Care in Cancer…
- 6 Semantic Scholar
LLM-driven rapid prototyping of microfluidic devices: a deterministic Python-mediated framework for design, fabrication, and validation
Microfluidic device design remains time-intensive and heavily reliant on computer-aided design (CAD) expertise, which limits scalability in rapid prototyping, h…
- 7 Semantic Scholar
The impacts of AI on new quality productive forces: an empirical study
The impacts of AI on new quality productive forces: an empirical study. Management & Marketing…
- 8 Semantic Scholar
Development of a Framework for Evaluating Large Language Model Safety and Reliability: a Proof-of-Concept Evaluation
Development of a Framework for Evaluating Large Language Model Safety and Reliability: a Proof-of-Concept Evaluation. Journal of medical systems…
- 9 Semantic Scholar
Evaluation of an AI-Powered Patient Education Application in Bariatric Surgery: A Prospective Mixed-Methods Feasibility Study
Evaluation of an AI-Powered Patient Education Application in Bariatric Surgery: A Prospective Mixed-Methods Feasibility Study. medRxiv…
- 10 Semantic Scholar
MAST-LLM: A large language model-driven multimodal adaptive spatio-temporal framework for air traffic complexity prediction
MAST-LLM: A large language model-driven multimodal adaptive spatio-temporal framework for air traffic complexity prediction. Applied intelligence (Boston)…
- 11 Semantic Scholar
Large language models in perioperative anesthesia: a risk-based roadmap for safe clinical adoption
Large language models in perioperative anesthesia: a risk-based roadmap for safe clinical adoption. Anesthesiology and Perioperative Science…
- 12 Semantic Scholar
Scientific reliability of physics-informed foundation models for engineering: a critical review, evidence-based assessment framework, and future research roadmap
Scientific reliability of physics-informed foundation models for engineering: a critical review, evidence-based assessment framework, and future research roadma…
- 13 arXiv
φ-RIE: From Photorealistic Reconstruction to Interactive Environments
3D Gaussian Splatting (3DGS) can reconstruct a captured scene photorealistically, but the resulting representation does not by itself support physical interacti…
- 14 arXiv
HARMONY: Hierarchical Agentic Reasoning for MONocular Image-to-Scene Synthesis
Compositional 3D scene reconstruction has recently been explored from two directions: agentic reasoning that provides semantic understanding of spatial relation…
- 15 arXiv
DreamStream: Towards Policy-Oriented Generative Simulation for End-to-End Driving
Faithfully evaluating end-to-end driving policies in simulation requires observations that are not merely photo-realistic, but preserve the scene features a pol…
- 16 arXiv
A Decentralized Partially Observable Team Decision Methodology with Delayed Information Sharing
We study decentralized partially observable team decision problems with low-rank latent dynamics and unknown system models. The proposed framework combines team…
- 17 arXiv
SpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue
Long-term conversational memory in multi-party settings requires more than retrieving relevant content from long-term conversations: it must distinguish who sai…
- 18 arXiv
CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agents
Agents often work on complex problems that require millions of tokens of context, which necessitates compacting across sessions due to limited context windows. …
- 19 arXiv
SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving
We introduce SWE-Serve, a benchmark for evaluating agents on production inference engineering tasks. Implementing an inference feature can require coordinating …
- 20 arXiv
A2M: Trace-Optimized Agent Hijacking in the MCP Ecosystem
Agents using the Model Context Protocol (MCP) rely on semantic matching to select tools from third-party servers, exposing a semantic supply-chain risk through …
- 21 arXiv
Grow the Harness, Not the Context: From Strategy-Free Scaffolds to Reusable Specialist Agents
Large language model (LLM) agents often handle streams of related tasks, yet standard harnesses repeatedly ask the model to reconstruct the same control decisio…
- 22 arXiv
Type-Safe Is Not Error-Free: A Constrained Decision Head Follows the Option Name, Not the Rubric Bound to It
Typed decision models are built for settings where model outputs are consumed directly by software. Instead of generating free-form text, they return a decision…
- 23 arXiv
FleXray: Universal Clinical X-ray Segmentation
X-ray is medicine's most widely used imaging modality, yet remains among its least quantitative. Unlike volumetric modalities like CT or MRI, X-ray collapses 3D…
- 24 arXiv
EquivSVA: A Formally Verified Dataset of Behavioral Assertions Across Equivalent RTL Implementations
Large language models are increasingly used to generate SystemVerilog Assertions from natural-language specifica- tions and register-transfer-level designs. Exi…
- 25 arXiv
Metrics Failure in LLM-Based Code Vulnerability Repair: An Empirical Study and a Change-Aware Screen
Large language models (LLMs) are increasingly applied to the automated repair of C/C++ security vulnerabilities, and compile rate is a commonly reported proxy f…
- 26 arXiv
Automatic depth-based local center clustering via $β$-integrated local depth and adaptive grouping
Clustering is an unsupervised learning technique that partitions unlabeled data into groups. Most existing methods require user-specified parameters, such as th…
- 27 arXiv
Diffusion-Induced Spatial Attention Overlapping Community Detection
Detection of overlapping communities is essential for modelling networks in which nodes participate simultaneously in multiple structural or functional groups. …
- 28 arXiv
Evaluating the Semantic-to-Geometric Gap in Adversarial Defenses Against Vision-Language Model-Based Plagiarism
The rapidly advancing capabilities of vision-language models (VLMs) present a systemic challenge to academic integrity. VLMs now allow students to bypass meanin…
- 29 arXiv
ASTRA-SR: Atmospheric Seeing and Turbulence Restoration for Astronomical Image Super-Resolution
Ground-based planetary imaging suffers from atmospheric turbulence, sensor noise, and limited sampling, making restoration a joint denoising, deblurring, and su…
- 30 arXiv
GAD-MambaUNet: Direction-Group Mamba with Gradient-Adaptive DINOv3 Distillation for Lightweight Medical Image Segmentation
In this paper, we proposed GAD-MambaUNet, a lightweight medical image segmentation network that combines efficient local modeling, direction--group state-space …
2026-09-22 · 週二 10 篇
- 1 👍 199
Realtime-Venus: A full-duplex interaction system with asynchronous delegation
A proactive full-duplex system with separate audio-visual and audio models enables continuous perception, native speech generation, and asynchronous tool execut…
- 2 👍 176
RRSI: Regularized Recursive Self-Improvement of Agent Harnesses
An LLM agent's capability is largely magnified by its harness, namely the prompts, control flow, tooling, memory, and context management surrounding the frozen …
- 3 👍 142
OmniEdu: Open Foundation Models for Learning and Teaching
Educational foundation models must solve problems, understand curriculum structure, diagnose learner difficulties, and provide appropriate instructional support…
- 4 👍 125
WorldCrafter: Consistent Video World Model with Implicit 3D-aware Memory
Video world models enable interactive exploration of dynamic environments, yet struggle to respect prior observations over long horizons and across viewpoints. …
- 5 👍 115
GameHorizon Suite: Multi-Horizon Data and Evaluation in Gameplay
Modern video games provide a measurable testbed for AI models, combining abilities of visual understanding, instruction decomposition, goal planning, and precis…
- 6 👍 104
Transferring the Intelligence of VLMs to Robotic Control
Humans can seamlessly adapt to both physical and digital worlds, suggesting that while a digital-to-real gap exists in embodiment, environment and task, human i…
- 7 👍 75
Grounded Action Model: 3D Grounding as a Foundation for Robotics
Manipulation policies must know which objects matter and where they are, yet the pretrained backbones that current robot foundation models build on, from langua…
- 8 👍 52
Document Retrieval-Aware Chunking (D-RAC): Universal Retrieval-Aware Ingestion of Enterprise Documents via PDF Normalization and Multimodal Markdown Conversion
Retrieval-Augmented Generation (RAG) systems over enterprise knowledge bases must ingest heterogeneous document formats -- PDFs, Word documents, presentations, …
- 9 👍 40
VideoGen-Agent: Reinforcing Video Generation Agents
Recent advances in video generative models have enabled high-fidelity, temporally coherent video generation. However, these models often struggle to satisfy pro…
- 10 👍 35
onPanda: Efficient Annotation of On-Policy Alignment Data for LLMs and Agents via Token-Level Correction
We present onPanda, an interactive tool for efficiently annotating LLM alignment data and agent trajectories. onPanda adopts token-level correction as its core …
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
AI information literacy in knowledge work: a pathway from tool awareness to responsible information practice
As generative artificial intelligence (GenAI) increasingly mediates information-intensive activities, greater attention is needed to the factors associated with…
- 2 Semantic Scholar
Tongue-pulse synergistic prescription recommendation via topology reconstruction and multi-source alignment in heterogeneous graphs
Tongue-pulse synergistic prescription recommendation via topology reconstruction and multi-source alignment in heterogeneous graphs. Network Modeling Analysis i…
- 3 Semantic Scholar
Structured Prompt and Agent AI Framework for Reliable LLM-Based RC Beam Design
Structured Prompt and Agent AI Framework for Reliable LLM-Based RC Beam Design. International Journal of Concrete Structures and Materials…
- 4 Semantic Scholar
Explainable optimized deep learning and generative AI based framework for finger millet disease detection in smart agriculture
Explainable optimized deep learning and generative AI based framework for finger millet disease detection in smart agriculture. Discover Artificial Intelligence…
- 5 Semantic Scholar
BootDiffLights: Decentralized traffic signal control via diffusion-based reward generation
BootDiffLights: Decentralized traffic signal control via diffusion-based reward generation. Applied intelligence (Boston)…
- 6 Semantic Scholar
RadCoT: a Radiological Chain-of-Thought framework for enhanced error detection in radiology reports
RadCoT: a Radiological Chain-of-Thought framework for enhanced error detection in radiology reports. European Radiology Experimental…
- 7 Semantic Scholar
Artificial intelligence in pediatric diabetes: clinical benefits, developmental risks, and ethical challenges
Artificial intelligence in pediatric diabetes: clinical benefits, developmental risks, and ethical challenges. European Journal of Pediatrics…
- 8 Semantic Scholar
A comprehensive survey of Edge-Fog-Cloud-IoT architectures: Intelligence, Security, and Sustainability
A comprehensive survey of Edge-Fog-Cloud-IoT architectures: Intelligence, Security, and Sustainability. Journal of King Saud University: Computer and Informatio…
- 9 Semantic Scholar
Fastvsearch: vision-language retrieval with object-level indexing for low-latency large-scale video search
Fastvsearch: vision-language retrieval with object-level indexing for low-latency large-scale video search. Machine Vision and Applications…
- 10 Semantic Scholar
ProtoSurv: prototype-guided adaptation of computed tomography foundation model for lung-cancer prognosis prediction
ProtoSurv: prototype-guided adaptation of computed tomography foundation model for lung-cancer prognosis prediction. Visual Computing for Industry, Biomedicine,…
- 11 Semantic Scholar
Uncoded Clinical Features from Multilingual Electronic Health Records in Catalonia: Development and Validation Study
Uncoded Clinical Features from Multilingual Electronic Health Records in Catalonia: Development and Validation Study. medRxiv…
- 12 Semantic Scholar
Large language models for teaching and learning in higher education: opportunities, challenges, and future directions
Large language models for teaching and learning in higher education: opportunities, challenges, and future directions. International Journal of Educational Tech…
- 13 Semantic Scholar
Cortex: A Retrieval-Augmented Framework for Career-Aligned Learning Paths
Cortex: A Retrieval-Augmented Framework for Career-Aligned Learning Paths. Machine-mediated learning…
- 14 Semantic Scholar
Preparing AI chatbots to respond to patient distress and suicidality in high-risk healthcare settings
Preparing AI chatbots to respond to patient distress and suicidality in high-risk healthcare settings. npj Digital Medicine…
- 15 Semantic Scholar
Artificial intelligence driven digital twin for dynamic reconfigurable intelligent surface control in terahertz internet of things mesh networks
Artificial intelligence driven digital twin for dynamic reconfigurable intelligent surface control in terahertz internet of things mesh networks. Discover Artif…
- 16 Semantic Scholar
A systematic review and Zero Trust governance framework for agentic UAV robotics in public safety
A systematic review and Zero Trust governance framework for agentic UAV robotics in public safety. Discover Robotics…
- 17 Semantic Scholar
Scalable Causal-Interpretable Machine Learning for Cancer Prescreening Using Electronic Health Records
Scalable Causal-Interpretable Machine Learning for Cancer Prescreening Using Electronic Health Records. medRxiv…
- 18 Semantic Scholar
Security smell detection in Dockerfiles
Security smell detection in Dockerfiles. International Conference on Automated Software Engineering…
- 19 Semantic Scholar
Computational Approaches to Penicillin Allergy De-labelling: A Clinical Review
Computational Approaches to Penicillin Allergy De-labelling: A Clinical Review. Clinical reviews in allergy and immunology…
- 20 Semantic Scholar
Iterative Demand Response Algorithm for Human-Computer Cooperation Integrating Knowledge of Large Language Model
Iterative Demand Response Algorithm for Human-Computer Cooperation Integrating Knowledge of Large Language Model. Applied intelligence (Boston)…
- 21 Semantic Scholar
From transformer to large language models for automated generation of hospital discharge summaries: a systematic review
From transformer to large language models for automated generation of hospital discharge summaries: a systematic review. Network Modeling Analysis in Health Inf…
- 22 Semantic Scholar
Beyond the gold standard: evaluating AI for subject indexing of Swedish LGBTQ + fiction
This study evaluates the potential of large language models (LLMs) for subject indexing of Swedish LGBTQ + fiction using the dedicated QLIT thesaurus, based on …
- 23 arXiv
Critical-State RL: Diagnosing Trainable States for Multi-Turn Tool Use
Multi-turn tool-use failures can hinge on a single model call, yet reward variation alone does not reveal which call would benefit from training. When rewards d…
- 24 arXiv
GAE: Learning a Geometry-Native Latent Space for 3D-Consistent World Generation
We present a compact geometry-native latent space as a shared foundation for perception and generation. Visual generators can produce photorealistic frames with…
- 25 arXiv
LoRA-generating hypernetworks for efficient on-device LLM generative personalization
On-device large language models (`LLMs'), e.g. running on mobile phones, are ripe for improvement via personalization. The limited compute resources of mobile d…
- 26 arXiv
DexTacWAM: A Visuo-Tactile World-Action Model for Dexterous Manipulation
Dexterous manipulation depends on contact dynamics that are often only partially observable from vision. Recent World-Action Models (WAMs) couple predictive vid…
- 27 arXiv
DolphinBench: Mapping the Pareto Frontier of Agent Memory
Agents today often take real-world actions that depend on long-term memory and context recall over time. However, most current memory benchmarks are built for a…
- 28 arXiv
Rare Event Estimation via Iterative Unalignment
As agents are deployed with increased autonomy, even extremely rare events along their stochastic output trajectories can occur and prove catastrophic. Safe dep…
- 29 arXiv
Emergent Collusion in Long-Horizon LLM Agent Interaction
LLM agents are increasingly deployed in collaborative settings, yet long-term interaction may give rise to undesirable coordination. We study the emergence of c…
- 30 arXiv
Jev for Scientific Decisions: Evaluating Semantic Choices and Their Consequences
Scientific workflows often require choosing among known relations before a deterministic calculation can proceed. Whether observations share a culture, treatmen…
2026-09-21 · 週一 10 篇
- 1 👍 106
CodeMidas: Scaling Agentic Coding RL Environments from Code Itself
Training capable coding agents via reinforcement learning (RL) requires diverse tasks with reliable verifiers. Open-source codebases offer a rich source of such…
- 2 👍 92
Grounded Skill Synthesis from Code at Scale for Agentic Intelligence
Code2Skill automatically extracts verifiable procedural skills from source code to improve agent performance before interaction experience accumulates.
- 3 👍 64
RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents
Computer-use agents (CUAs) have advanced along two separate lines: graphical interaction and software development through code and the command line. Real digita…
- 4 👍 37
IntBMoE: Integrating Block-Level Conditioning into Expert Composition for Full-Participation Mixture-of-Experts
Mixture-of-Experts (MoE) scales capacity, but existing designs cannot set three quantities independently. For a single token, participation is how many experts …
- 5 👍 31
OmniVChat: Synthesizing, Benchmarking, and Training for Native Audio-Visual Dialogue
We define OmniVChat (Omni Video Chat) as the task of native audio-visual dialogue between a user and an omni model. In OmniVChat, omni models directly and simul…
- 6 👍 25
Designer-RSI: Evolving Procedural Memory from User Traffic for Agentic Graphic Design
Professional graphic design is a long-horizon agentic task in which structured, editable artifacts emerge from many interdependent actions, yet outcomes admit n…
- 7 👍 20
OmniVBench: A Benchmark and Large-Scale Dataset for Omni Reference-to-Video Generation
Reference-to-video (R2V) generation is evolving toward increasingly general and versatile reference control, giving rise to the emerging paradigm of omni R2V ge…
- 8 👍 15
BI-Agent and BI-Bench: Towards Automating End-to-End Business Intelligence
Business intelligence (BI) is a cornerstone of enterprise decision-making and is widely used by enterprise users in software such as Power BI and Tableau. In tr…
- 9 👍 14
MintAct: A Unified Visual Agent for Digital Environments
We present MintAct, a family of vision-language models that unifies UI grounding, multi-step navigation across mobile, desktop, and web, and visual tool use, tr…
- 10 👍 9
GraphSkillEvo: Evolutionary Optimization of Graph-Structured Agent Skills
Skills can improve the performance of Large Language Model (LLM) agents by providing task-specific procedural guidance, while skill optimization further improve…
更多論文（arXiv / ChatPaper / Semantic Scholar，26 篇）
- 1 Semantic Scholar
Developing nonviolent artificial intelligence: nonviolence as a principle of ethical AI
Developing nonviolent artificial intelligence: nonviolence as a principle of ethical AI. AI and Ethics…
- 2 Semantic Scholar
Seeing the subtle: lesion-aware multi-scale fusion for scribble-supervised segmentation
Seeing the subtle: lesion-aware multi-scale fusion for scribble-supervised segmentation. Multimedia Systems…
- 3 Semantic Scholar
The ramifications of LLM watermarks for authorship and groups
The ramifications of LLM watermarks for authorship and groups. AI and Ethics…
- 4 Semantic Scholar
Methodological transitions in mental health research driven by machine learning
Methodological transitions in mental health research driven by machine learning. BJPsych Open…
- 5 Semantic Scholar
S-AFCM: Stabilizing Adaptive Feature Capture Method via Local Density Smoothing and Diversity-Enhanced Sampling
S-AFCM: Stabilizing Adaptive Feature Capture Method via Local Density Smoothing and Diversity-Enhanced Sampling. Journal of Mathematical Imaging and Vision…
- 6 Semantic Scholar
Probing context-sensitive expectations through user-driven personality configuration in conversational AI
Abstract Conversational agents are increasingly expected to adapt across contexts and evolve their personalities through interactions, yet most remain static on…
- 7 Semantic Scholar
Learning Based Parameter Estimation in Diffusion Models-An Identifiability Study
This work proposes a learning-based biosensing receiver (BSR) for estimating diffusion parameters in molecular communication (MC) systems. It is done from timel…
- 8 Semantic Scholar
Class balanced diabetic retinopathy image synthesis using a latent diffusion framework
Class balanced diabetic retinopathy image synthesis using a latent diffusion framework. Discover Artificial Intelligence…
- 9 Semantic Scholar
GLoT: global–local token alignment for interpretable zero-shot graph reasoning with LLMs for e-commerce applications
GLoT: global–local token alignment for interpretable zero-shot graph reasoning with LLMs for e-commerce applications. Multimedia Systems…
- 10 Semantic Scholar
From access to engagement: AI literacy, technological play and academic libraries as sociotechnical learning environments
This study examines how academic libraries can support AI literacy through technological play, informal learning, librarian mediation and technology-rich spaces…
- 11 Semantic Scholar
Communicating academic integrity policy in the age of generative AI
This study aims to examine how students and teaching staff perceive and enact academic-integrity policies in Romanian universities under the normative uncertain…
- 12 Semantic Scholar
Diabetes question answering method based on knowledge graph and large language model
Diabetes question answering method based on knowledge graph and large language model. Discover Artificial Intelligence…
- 13 Semantic Scholar
A Multidimensional Assessment of AI Literacy Services in University Libraries
Abstract As artificial intelligence (AI) continues to reshape the landscape of higher education, academic libraries are emerging as strategic actors in advancin…
- 14 Semantic Scholar
Evidence-grounded multimodal mining of fine-grained research problems and methods from computer vision papers
Evidence-grounded multimodal mining of fine-grained research problems and methods from computer vision papers. Multimedia Systems…
- 15 Semantic Scholar
Artificial intelligence in financial risk management empirical evidence from Morocco and North Africa
Artificial intelligence in financial risk management empirical evidence from Morocco and North Africa. Discover Artificial Intelligence…
- 16 Semantic Scholar
Mapping the role of artificial intelligence in renewable energy and the clean energy transition: a bibliometric review
This study aims to present a bibliometric analysis of research on artificial intelligence (AI) applications in renewable energy (RE) technologies and the clean …
- 17 Semantic Scholar
Controllable flow fields for high-fidelity virtual try-on with diffusion models
Controllable flow fields for high-fidelity virtual try-on with diffusion models. Multimedia Systems…
- 18 Semantic Scholar
AI supported doodling and gallery walk defence provide visible evidence of student reasoning in higher education assessment
AI supported doodling and gallery walk defence provide visible evidence of student reasoning in higher education assessment. Discover Artificial Intelligence…
- 19 Semantic Scholar
Design and implementation of an end to end music audio generation system based on diffusion probabilistic model
Design and implementation of an end to end music audio generation system based on diffusion probabilistic model. Discover Artificial Intelligence…
- 20 Semantic Scholar
Latency-efficient cloud-edge cultural pattern generation via lossless lightweight condition transmission and bandwidth-adaptive scheduling
Latency-efficient cloud-edge cultural pattern generation via lossless lightweight condition transmission and bandwidth-adaptive scheduling. Discover Artificial …
- 21 Semantic Scholar
Artificial intelligence data centers could reach one percent of global electricity demand by 2030
Artificial intelligence data centers could reach one percent of global electricity demand by 2030. Communications Sustainability…
- 22 Semantic Scholar
On the Role of Relay Cells in Information Propagation Speed for Dictyostelium Chemotaxis: An Agent-Based Modeling Study
Chemotactic signaling in Dictyostelium discoideums (DDs) relies on the secretion, diffusion, and enzymatic degradation of cyclic adenosine monophosphat (cAMP) m…
- 23 Semantic Scholar
When AI summarizes the community: how generative AI summaries shape users’ intention to engage with original content
The integration of Generative Artificial Intelligence (GenAI) into user-generated content (UGC) communities has fundamentally reshaped how users seek and consum…
- 24 Semantic Scholar
A hybrid transformer with spatiotemporal attention for citrus pest and disease forecasting
A hybrid transformer with spatiotemporal attention for citrus pest and disease forecasting. Discover Artificial Intelligence…
- 25 Semantic Scholar
LLM-enabled Natural History Study Analysis to Support Rare Disease Research
LLM-enabled Natural History Study Analysis to Support Rare Disease Research. medRxiv…
- 26 Semantic Scholar
Leveraging Large Language Models for Colorectal Cancer Symptom Extraction from MIMIC-IV Clinical Notes
Leveraging Large Language Models for Colorectal Cancer Symptom Extraction from MIMIC-IV Clinical Notes. medRxiv…
2026-09-20 · 週日 0 篇
更多論文（arXiv / ChatPaper / Semantic Scholar，27 篇）
- 1 Semantic Scholar 📖 1
Exploring Veo 3's Capabilities for Generating Urban Traffic Scenes in 76 Cities Worldwide
This study explores the potential of Google Veo 3, a generative video model, to synthesise 8-second dashcam-style urban traffic scenes based solely on text prom…
- 2 Semantic Scholar
AI-based diagnosis of the continuity of research thought: A case study from the perspective of innovation and management accounting
<p style="text-align:justify"><span style="font-size:11pt"><span style="background:white"><span style="line-height:105%"><span style="font-family:"Times New Rom…
- 3 Semantic Scholar
Memristive singular value decomposition
Memristive singular value decomposition. Nature Communications…
- 4 Semantic Scholar
A survey on knowledge graph-augmented large language model reasoning: theoretical challenges, technical pathways, and evolutionary logic
A survey on knowledge graph-augmented large language model reasoning: theoretical challenges, technical pathways, and evolutionary logic. Knowledge and Informat…
- 5 Semantic Scholar
Patterns of AI-generated teaching recommendations in accounting education: A qualitative analysis of ChatGPT responses
<p style="text-align:justify"><span style="font-size:11pt"><span style="line-height:105%"><span style="font-family:"Times New Roman",serif"><span style="letter-…
- 6 Semantic Scholar
Student Research Track: Hack on the Road: Can LLM-Based Driving Agents Improve Driver Response During Automated Vehicle Cyber-Attacks?
Automated vehicles (AVs) increasingly rely on connected communication systems, which make them vulnerable to cyber-attacks. During such events, drivers may not …
- 7 Semantic Scholar
Student Research Track: Letting the Automated Vehicle Explain Its Own Errors: A Simulator Testbed for Passenger-Facing Explanation HMIs in SOTIF Error Situations
A fully automated (SAE L5) vehicle can obey every traffic rule and still behave near-erroneously, and a passenger who cannot tell why loses situation awareness …
- 8 Semantic Scholar
Sim-DSE: Mediating Multi-User Automations in Cars through Simulation-Augmented Decision Space Exploration
In-vehicle interaction is often envisioned as requiring as little explicit user input as possible. In multi-user scenarios, however, varying user states and exp…
- 9 Semantic Scholar
Dialog with trucks - A driving simulator study exploring situationally-aware conversational agents
Current AI-based conversational agents (CAs) largely lack situational awareness, limiting their ability to understand driving context, driver state, and immedia…
- 10 Semantic Scholar
The Role of Conversational Agents for Long-haul Truck Drivers: An Exploratory On-road Study
While conversational AI agents (CAs) have the potential to support drivers through hands-free and eyes-free interaction, their role in long-haul trucking remain…
- 11 Semantic Scholar
"Do as I say": Analysis of In-vehicle Assistant Command Pattern Differences between Germany and Mexico
In-vehicle assistants (IVAs) are moving from fixed command grammars toward large language model (LLM) driven conversational agents, where the quality of the sys…
- 12 Semantic Scholar
Exploring the Design Space of Adaptive Explanations in Automated Vehicles
Explanations in automated vehicles have been proposed to support users’ understanding, trust, and acceptance of automated driving behavior. However, many explan…
- 13 Semantic Scholar
Interviewing Silicon Experts: A Persona-Based LLM Interview Pipeline in Automated Driving
Human expert interviews are valuable but often limited by slow recruitment, scarce expertise, and network-based sampling, especially for abstract human-automati…
- 14 Semantic Scholar
From Assistance to Affective Co-Driving: Novice Drivers’ Emotional Needs and Design Expectations for In-Vehicle AI Agents
As AI becomes increasingly embedded in intelligent cockpits, human-vehicle interaction is shifting from task assistance toward affective and collaborative drivi…
- 15 Semantic Scholar
Artificial Intelligence to Support Clinical Practice Across the Inpatient Heart Failure Care Pathway: A Systematic Review
Heart failure is a complex clinical syndrome requiring repeated reassessment, close monitoring, and timely treatment decisions throughout hospitalization. With …
- 16 Semantic Scholar
Constrained Generative UI for the Cockpit: A Comparison of LLM-Assembled and Pre-Defined Interfaces
In-car infotainment systems are growing in complexity, making traditional design and engineering approaches costly and difficult to scale. Recent advances in la…
- 17 Semantic Scholar
Personalised Electric Vehicle Acoustics with Generative AI: A Dynamic Sonification and User Acceptance Study
Electric vehicles (EVs) reduce powertrain noise, creating safety challenges and opportunities for sound design. This paper examines whether generative audio sup…
- 18 Semantic Scholar
Emotune: Exploring Context-Aware LLM Support for Driver Emotion Regulation
Driver emotions such as stress, frustration, and anxiety are associated with riskier behavior and reduced situational awareness, motivating in-vehicle systems t…
- 19 Semantic Scholar
Role of foundation models in data-driven tissue diagnostics.
Artificial intelligence (AI)-based tissue diagnostics is entering a new phase driven by pathology foundation models: large-scale encoders and multimodal systems…
- 20 Semantic Scholar
Artificial intelligence in science education for SDG 4: A systematic review of AI interventions for ESD and 21st-century competencies
Artificial intelligence (AI) is transforming science education and creating new opportunities to advance education for sustainable development. This systematic …
- 21 Semantic Scholar
Privacy-Constrained Distributionally Robust Detection and Collaborative Attribution of LLM-Generated Text
With the widespread use of large language models, distinguishing human-written text from LLM-generated text has become increasingly important for content authen…
- 22 Semantic Scholar
Transformer-Based Multimodal Image Fusion in Healthcare: A Comprehensive Survey of Architectures, Fusion Strategies, and Clinical Applications
Transformer architectures have changed how multimodal medical image fusion works. Convolutional neural network (CNN)-based methods are constrained to local rece…
- 23 Semantic Scholar
Uncovering Latent Scientific Structures in Albanian Texts Using Transformer-Based Clustering
This study analyzes the semantic structure of an Albanian corpus comprising scientific texts with 11.5 million words and approximately 74,500 paragraphs. It aim…
- 24 Semantic Scholar
AI-Enabled Prescribing in Opioid Use Disorder Care: Decision Support, Risk Scores, Large Language Model Copilots, and Digital Therapeutics (2024-2025).
AI-Enabled Prescribing in Opioid Use Disorder Care: Decision Support, Risk Scores, Large Language Model Copilots, and Digital Therapeutics (2024-2025).. Europea…
- 25 Semantic Scholar
Cybersecurity and AI: A Comprehensive Implementation of Advanced Large Language Models for Threat Detection, Digital Forensics, SOC Automation, and Security Vulnerability Mitigation
This report encompasses the cutting-edge implementation of advanced Large Language Models (LLMs) for cybersecurity and digital forensics applications at the int…
- 26 Semantic Scholar
Beyond LLMs: A Post-Transformer World Emerges
A focus of the next decade might be on teaching AI how to master reality and understand how things work in the physical world.…
- 27 Semantic Scholar
Automated Validation in Embedded IoT Development: A Review of Continuous Integration and Delivery across the Fidelity Ladder
Embedded software for the Internet of Things is developed under pressures that pull in opposite directions: the delivery cadence of connected products, sustaine…
2026-09-19 · 週六 0 篇
更多論文（arXiv / ChatPaper / Semantic Scholar，6 篇）
- 1 Semantic Scholar 📖 1
scBaseCount: An AI agent-curated, standardized, auto-updated single-cell data repository.
Single-cell RNA sequencing has transformed cell biology by enabling precise transcriptomic measurements of individual cells. The Sequence Read Archive (SRA) is …
- 2 Semantic Scholar 📖 1
MCSched: Memory-Controller-Aware Scheduling for Embodied LLM Workloads on NVIDIA Jetson
MCSched: Memory-Controller-Aware Scheduling for Embodied LLM Workloads on NVIDIA Jetson.…
- 3 Semantic Scholar
Artificial intelligence in scientific research: a practical guide to tools and ethical considerations
Artificial intelligence in scientific research: a practical guide to tools and ethical considerations. Egyptian Journal of Bronchology…
- 4 Semantic Scholar
Using Multimodal Large Language Models for False Alarm Reduction in Image-Based Fire Detection
Using Multimodal Large Language Models for False Alarm Reduction in Image-Based Fire Detection. Fire technology…
- 5 Semantic Scholar
Prospecting the protein design landscape.
Prospecting the protein design landscape.. FEBS Letters…
- 6 Semantic Scholar
Avoiding common failures in AI for health and medicine.
The widespread adoption of artificial intelligence (AI) in healthcare necessitates reliable AI systems. Reliability refers to clinically acceptable performance …
2026-09-18 · 週五 10 篇
- 1 👍 93
DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
The widespread adoption of long-horizon agents has made model workloads increasingly input-heavy. Although prior work has substantially reduced the cost of long…
- 2 👍 77
Can MiniMax-H3 Reason About the Physical World? An Evaluation of Omni-Modal Generative Model
Recent Omni-Modal Generative Models (Omni-Models) have advanced content generation toward unified modeling of text, images, video, and audio. MiniMax-H3 exempli…
- 3 👍 74
When EOS Tokens Disagree: Understanding Length Inflation in On-Policy Distillation
We study length inflation in on-policy distillation (OPD), where student responses can become excessively long and even exhaust the generation budget. We identi…
- 4 👍 68
SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness
As coding agents move from supervised code completion to unattended, around-the-clock exploration, their work expands from isolated predictions into long trajec…
- 5 👍 55
An Empirical Study of Harness Design for Coding Agents
Coding harnesses shape how autonomous coding agents translate model capabilities into long-horizon software-engineering performance, yet existing work typically…
- 6 👍 38
JEPA-Anything: Learning Predictive Models across Different Worlds
World modeling enables intelligence to anticipate consequences, guide interventions, and learn from interaction. Yet predictive models remain domain-specific: c…
- 7 👍 28
RiskChainBench: A Benchmark for Obfuscated Platform Message Restoration and Evidence-Grounded Web Investigation
Platform abuse campaigns conceal redirection instructions with emojis, homophones, character decomposition, and redundant symbols, then route users through disg…
- 8 👍 26
RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning
Multi-turn agents trained with reinforcement learning (RL) receive a single scalar reward per trajectory, which motivates self on-policy distillation (OPD) to s…
- 9 👍 25
Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents
GUI agents execute long-horizon tasks on dynamic graphical user interfaces, where pop-ups, delayed loads, and relocated widgets routinely invalidate plans fixed…
- 10 👍 25
Self-Evolving Search Index
Information retrieval is increasingly important as LLM agents tackle complex tasks involving diverse information needs. Because retrieval relies on an index tha…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
From obligation to enforcement: mapping EU AI act and CRA cybersecurity requirements to technical controls for LLM-based autonomous agents
From obligation to enforcement: mapping EU AI act and CRA cybersecurity requirements to technical controls for LLM-based autonomous agents. International Journa…
- 2 Semantic Scholar
Counterfactual reasoning and causal intervention for multi-modal rumor detection
Counterfactual reasoning and causal intervention for multi-modal rumor detection. Knowledge and Information Systems…
- 3 Semantic Scholar
An Analysis of the Combined Effects of Advanced Information Technologies and Enterprise Risk Management on Organizational Performance
This study aims to investigate the combined effects of advanced information technologies that include data analytics, enterprise resource planning (ERP) and the…
- 4 Semantic Scholar
A population representative XAI framework for transparent dysglycemia screening in Indonesia
A population representative XAI framework for transparent dysglycemia screening in Indonesia. Discover Artificial Intelligence…
- 5 Semantic Scholar
Algorithmic Fairness, Meritocracy, and Institutional Justice
Algorithmic Fairness, Meritocracy, and Institutional Justice. Philosophy & Technology…
- 6 Semantic Scholar
Multimodal content generation and enhancement for smart imaging
Multimodal content generation and enhancement for smart imaging. Discover Artificial Intelligence…
- 7 Semantic Scholar
The pitfalls of AI detection in academic writing: bias, false positives, and the need for inclusive assessment
The pitfalls of AI detection in academic writing: bias, false positives, and the need for inclusive assessment. AI and Ethics…
- 8 Semantic Scholar
Erratum to: LLM-enhanced-EA-driven automated optimization framework for joint indoor deployment of multiple BSs and RISs
Erratum to: LLM-enhanced-EA-driven automated optimization framework for joint indoor deployment of multiple BSs and RISs. Science China Information Sciences…
- 9 Semantic Scholar
Enhancing vocabulary learning in EAP: investigating AI-supported spaced retrieval, student experiences, and perceptions
Enhancing vocabulary learning in EAP: investigating AI-supported spaced retrieval, student experiences, and perceptions.…
- 10 Semantic Scholar
Towards Understanding Token Evolution in Continual Learning of MLLMs
Towards Understanding Token Evolution in Continual Learning of MLLMs. International Journal of Computer Vision…
- 11 Semantic Scholar
Generative AI in peer review: evolving policies and regulatory pathways for scientific rigour
Generative AI in peer review: evolving policies and regulatory pathways for scientific rigour. AI and Ethics…
- 12 Semantic Scholar
Interaction of network and customer characteristics on product diffusion on social media platforms: a hybrid multi-agent model
Interaction of network and customer characteristics on product diffusion on social media platforms: a hybrid multi-agent model. Computational and mathematical o…
- 13 arXiv
Coding Agents with an Obstacle-Aware Harness for Safe Robot Manipulation
Coding agents have emerged as a promising paradigm for robot manipulation: a language model writes the robot controller as a program, and agents built in this w…
- 14 arXiv
Embedding Models Measure in Peculiar Ways
Embedding spaces define notions of semantic similarity and distance. We study whether those embeddings reflect physical measurements of mass, distance, time and…
- 15 arXiv
Workspace Models: Lightweight Robotic Memory via Saliency-Driven Supervision
Complex robotic manipulation tasks frequently require a long-term memory of past events and actions. As conditioning on full histories renders policies prone to…
- 16 arXiv
Can 4D Foundation Models Remember?
Perceiving and remembering the visual world is fundamental to navigating and interacting with our environment. Current 4D foundation models, such as camera-cont…
- 17 arXiv
SplashSplat: Reconstructing Splashing Liquids from Real-World Multi-View Videos
A splash lives for a fraction of a second: sheets tear into ligaments and droplets, appearance is view-dependent and nearly textureless, and little persists lon…
- 18 arXiv
Paint-Anything: Unified Any-Color Control for Image Generation and Editing
Professional design requires any-color control: the ability to specify an object's target color with any 24-bit hex value for image generation and editing. Prio…
- 19 arXiv
ERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis
Hereditary polyposis syndromes can be precursor lesions to colorectal cancer and are associated with a broad spectrum of extracolonic tumors. Early identificati…
- 20 arXiv
How Does Distribution Shift Shape Pretraining Gains in Neural PDE Surrogates?
Pretraining a neural PDE surrogate can reduce the amount of new CFD data needed when geometry or modeled physics changes. However, it remains unclear how differ…
- 21 arXiv
Quantifying Overclaiming Propensity in Frontier LLM Agents
Frontier coding agents are increasingly trusted to work autonomously for long periods, yet an agent's final response is often the only account of that work a us…
- 22 arXiv
Unifying Models of Intergroup Hostility in Online Discourse
Hostile rhetoric toward social groups can normalize exclusion and justify mistreatment, as well as contribute to rising polarization and political violence. Eff…
- 23 arXiv
Score Centering Stabilizes Off-policy Reinforcement Learning
Reinforcement learning (RL) of large language models is notoriously sensitive to small differences between training and inference engines, often referred to as …
- 24 arXiv
PosteriorBench: From Point Estimates to Posterior Matching in Evaluating Generative Inverse Solvers
Generative models are increasingly used to solve scientific inverse problems, but existing evaluations still focus primarily on whether a method can produce a s…
- 25 arXiv
Harm Laundering in GPT Models: Evidence That Gender Discrimination Is Transformed Rather Than Reduced Across Safety-Trained Generations
Safety evaluations for large language models rely on surface-form classifiers that report declining harm scores across model generations. We provide evidence th…
- 26 arXiv
GeoAAC: Geometry-Based Adaptive Action Chunking from Denoising Trajectories in VLA Policies
Action chunking is widely used for action generation and execution in Vision-Language-Action (VLA) policies, yet existing approaches commonly use a fixed action…
- 27 arXiv
FlowSGS: Improving Flow Matching Priors for Inverse Imaging with Stochastic Interpolants
Flow matching has emerged as the state-of-the-art generative model and has been used for plug-and-play (PnP) priors to solve inverse problems in computational i…
- 28 arXiv
Semantic Action Graph: A Shared Representation for Agent Grounding and Human Interpretation of Sports Highlights
Generative agents are increasingly used to select and narrate video highlights, but they typically operate over unstructured or frame-level representations. The…
- 29 arXiv
Calibrated RF-Fingerprinting Under Interference With Heterogeneous Transmission Protocols
Radio Frequency(RF)-Fingerprinting is a spectrum monitoring technique that identifies specific transmitters based on hardware impairments imprinted within the e…
- 30 arXiv
Agile-WAM: An Agile Tactile World Action Model for Contact-Rich Robot Control
World Action Models (WAMs) advance beyond conventional visuomotor policies by jointly predicting future world states and robot actions, enabling the policy to l…
2026-09-17 · 週四 10 篇
- 1 👍 73
ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environments
Scientific code repositories encode decades of human knowledge in executable models, methods, and tools. Yet fragmented toolchains, implicit domain conventions,…
- 2 👍 62
Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening
In reinforcement learning for large language models, Proximal Policy Optimization (PPO) commonly uses a critic to estimate state values and reduce the variance …
- 3 👍 52
Confidence Comes from Experience: Experiential Confidence Estimation from Reasoning to Agents
Reliable confidence estimation is increasingly central to the trustworthy deployment of language models: a calibrated estimate of the probability that an output…
- 4 👍 46
ProgramDistill: From Interactive Web Apps to Verifiable Reference-Guided SWE Tasks
Coding agents are typically evaluated with desired behavior specified through issues or instructions. In practical web development, however, agents may need to …
- 5 👍 41
Agora: Git as Shared Memory for Collective AutoResearch
Autonomous research loops such as AutoResearch show that one coding agent can improve a training setup unattended. Run several of them and each session starts f…
- 6 👍 39
ActionPiece: Rethinking Action Tokenization for Autoregressive Vision-Language-Action Models
Action tokenizers play a central role in autoregressive vision-language-action (VLA) models, determining both the targets for policy training and the executable…
- 7 👍 31
EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents
Large language model (LLM) trading agents can combine market data, news, and executable analysis, but their behavior is often controlled by static hand-written …
- 8 👍 29
Zing-0.5: Toward Playable Worlds with Real-Time Joint Action and Text Control
We introduce Zing-0.5, a 5B autoregressive world model designed for playability: users can explore generated worlds, influence unfolding events, and respond to …
- 9 👍 23
SpectralShift: Effective Context Window Extension of Gated DeltaNet via Spectral Reparameterization
Recently, linear attention layers have been increasingly adopted to replace softmax attention at scale for long-context modeling. However, existing context exte…
- 10 👍 22
A Zeroth-Order Paradigm for LLM Preference Alignment
Direct preference alignment methods are widely used to align large language models (LLMs) with human preferences because of their computational and memory effic…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
A hybrid framework combining large language models and Shapley-based methods for explainable cyberbullying detection
A hybrid framework combining large language models and Shapley-based methods for explainable cyberbullying detection. Discover Artificial Intelligence…
- 2 Semantic Scholar
Enhancing recommendation systems with weighted aspect based opinion mining utilizing advanced deep learning for adaptive personalization of user preferences
Enhancing recommendation systems with weighted aspect based opinion mining utilizing advanced deep learning for adaptive personalization of user preferences. Wo…
- 3 Semantic Scholar
DiffDGCL: diffusion model and dynamic graph contrastive learning for multimodal recommendation
DiffDGCL: diffusion model and dynamic graph contrastive learning for multimodal recommendation. Multimedia Systems…
- 4 Semantic Scholar
Using the drift diffusion model to examine relationships between mental health and media selection dynamics
Using the drift diffusion model to examine relationships between mental health and media selection dynamics. Communication and Change…
- 5 Semantic Scholar
Cost efficient multimodal LLM deployment integrating tokenization economics MLOps and FinOps environments
Cost efficient multimodal LLM deployment integrating tokenization economics MLOps and FinOps environments. Discover Computing…
- 6 Semantic Scholar
Social learning within groups: Hypernetworks capture effects of within-group dynamics on social contagions
Social learning within groups: Hypernetworks capture effects of within-group dynamics on social contagions. Behavioral Ecology and Sociobiology…
- 7 Semantic Scholar
Adaptive large language model scaffolding supports online programming education through learning analytics and knowledge tracing
Adaptive large language model scaffolding supports online programming education through learning analytics and knowledge tracing. Discover Education…
- 8 Semantic Scholar
HydroSuite-AI: a web-based LLM environment for hydrological code generation and execution for the hydrosuite open-source ecosystem
HydroSuite-AI: a web-based LLM environment for hydrological code generation and execution for the hydrosuite open-source ecosystem. Earth Science Informatics…
- 9 Semantic Scholar
Interaction-induced knowledge narrowing risk in LLM systems
Interaction-induced knowledge narrowing risk in LLM systems. Discover Artificial Intelligence…
- 10 Semantic Scholar
Classifier chains for multi-label learning: a systematic review and taxonomy
Classifier chains for multi-label learning: a systematic review and taxonomy. Data mining and knowledge discovery…
- 11 Semantic Scholar
How generative AI transforms relational archives into simulations of living persons
How generative AI transforms relational archives into simulations of living persons. Discover Artificial Intelligence…
- 12 Semantic Scholar
Exploratory Implementation and Feasibility Report of CLASS (Clinical LLM Abstraction & Structuring System), A Large Language Model Pipeline for Extracting Unstructured Data From Clinical Notes
Exploratory Implementation and Feasibility Report of CLASS (Clinical LLM Abstraction & Structuring System), A Large Language Model Pipeline for Extracting Unstr…
- 13 Semantic Scholar
A Deployed Custom Fine-tuned Transformer Language Model Framework for Indian Fake News Detection
A Deployed Custom Fine-tuned Transformer Language Model Framework for Indian Fake News Detection. International Journal of Machine Learning and Cybernetics…
- 14 Semantic Scholar
FUS-DETR: a robust algorithm for drone detection from a cluttered background in airborne images
FUS-DETR: a robust algorithm for drone detection from a cluttered background in airborne images. Applied intelligence (Boston)…
- 15 Semantic Scholar
Multimodal large language models for bladder tumor detection in cystoscopy: a retrospective benchmarking study
Multimodal large language models for bladder tumor detection in cystoscopy: a retrospective benchmarking study. World journal of urology…
- 16 Semantic Scholar
Autonomous governance integrating agentic AI and zero trust for intelligent cybersecurity in distributed enterprise ecosystems
Autonomous governance integrating agentic AI and zero trust for intelligent cybersecurity in distributed enterprise ecosystems. Discover Internet of Things…
- 17 Semantic Scholar
LAMP: Long-Tailed Multi-modal Prompt Tuning for Vision-Language Models
LAMP: Long-Tailed Multi-modal Prompt Tuning for Vision-Language Models. International Journal of Computer Vision…
- 18 Semantic Scholar
Ethical integration of generative artificial intelligence in Islamic religious education involves human mediation and source verification
Ethical integration of generative artificial intelligence in Islamic religious education involves human mediation and source verification. Discover Artificial I…
- 19 Semantic Scholar
Between gatekeeping and coping: Iranian EFL students’ lived experience of AI-mediated writing assessment
Between gatekeeping and coping: Iranian EFL students’ lived experience of AI-mediated writing assessment. Language Testing in Asia…
- 20 Semantic Scholar
A retrieval-augmented model guided zero-shot prompt for LLM to detect mental disorders in social media text
A retrieval-augmented model guided zero-shot prompt for LLM to detect mental disorders in social media text. International Journal of Machine Learning and Cyber…
- 21 Semantic Scholar
Applications and challenges of artificial intelligence in cancer management
Applications and challenges of artificial intelligence in cancer management. Discover Applied Sciences…
- 22 Semantic Scholar
ObfXhate: a large-scale Arabic dataset and transformer-based framework for detecting obfuscated hate speech
ObfXhate: a large-scale Arabic dataset and transformer-based framework for detecting obfuscated hate speech. Language Resources and Evaluation…
- 23 Semantic Scholar
Whose voices count? Multilingualism, AI policy, and equity at Hispanic-Serving institutions
Whose voices count? Multilingualism, AI policy, and equity at Hispanic-Serving institutions. Language Policy…
- 24 Semantic Scholar
Explainable course recommendation with knowledge graphs: a comparative audit of diverse modeling paradigms
Explainable course recommendation with knowledge graphs: a comparative audit of diverse modeling paradigms. Data mining and knowledge discovery…
- 25 arXiv
Objective vs. Search: Decomposing What Makes a Good Tokeniser
Two dominant tokenisation algorithms are used by modern language models: byte-pair encoding (BPE) and UnigramLM. These differ along two orthogonal axes: their o…
- 26 arXiv
PointZero: 3D Point Track Completion for Learning Transferable 3D Dynamics
World models endow perceptual systems with the ability to predict how scenes evolve under interaction. They are most beneficial when trained on diverse volumes …
- 27 arXiv
Dreaming the Sound of Contact: Leveraging Video and Audio Generation for Zero-Shot Force-Aware Manipulation and Data Generation
Recent advances in video generation allow robots to learn manipulation trajectories from generated videos. However, these approaches produce purely kinematic tr…
- 28 arXiv
Exponential Hardness of Off-Policy Evaluation under History-Dependent Logging
Can a logged dataset visit every hidden state frequently and still be exponentially uninformative about a target policy's value? We show that it can when the lo…
- 29 arXiv
Cognitive Extensions for Dual-Process Language Agents: Memory and Self-Reflection in Interactive Environments
Language agents remain brittle in interactive environments, where success requires long-horizon state tracking, valid action execution, and recovery from failed…
- 30 arXiv
Affora: A Design System for Agent-Friendly Interfaces
Computer-use agents increasingly operate software designed for people, but interfaces often leave actions or task state unclear to machine readers. We present A…
2026-09-16 · 週三 10 篇
- 1 👍 294
Continual Learning Mechanisms Compose for Long-Horizon Memorization
Composing complementary continual learning mechanisms substantially improves long-horizon memorization in language models by combining data, function, and weigh…
- 2 👍 113
AI for Games in the Foundation Model Era
Foundation models, alongside advances in learned game-world models, are reshaping AI across the game lifecycle. Beyond playing games, recent systems model playe…
- 3 👍 99
StepAudio 3 Realtime Technical Report
Realtime spoken interaction demands deep reasoning, prompt responses, and fluid turn-taking. We present StepAudio 3 Realtime, an audio-language foundation model…
- 4 👍 75
StepAudio 3 Music Technical Report
We introduce StepAudio 3 Music, a large-scale, long-form music generation model that supports explicit musical planning and open-domain text-controlled generati…
- 5 👍 58
Generalized Agent Iteration: One Formal Framework for Iterative Policy Improvement and Recursive Self-Improvement
The paper proposes Generalized Agent Iteration as a unified formal framework for iterative policy improvement and recursive self-improvement, defining key axes …
- 6 👍 20
ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents
We introduce and release ScienceBuddy, an interactive scientific research workspace that brings continually improving scientific agents into researchers' everyd…
- 7 👍 16
HarnessVLN: Unifying Training-Free Embodied Navigation through an Agent Harness
Embodied navigation requires agents to interpret visual observations, accumulate spatial knowledge, and execute actions to follow instructions or locate objects…
- 8 👍 13
FLAT: Resampling Image and Text into 1D Flexible-Length Aligned Transmodal Tokens for Retrieval and Generation
Traditional multimodal representation learning and generation are two stages: a contrastive or self-supervised visual encoder is trained first, followed by a se…
- 9 👍 9
ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement
Recent work extends recursive self-improvement (RSI) to agent harnesses for long-horizon coding and terminal tasks, enabling agents to improve execution mechani…
- 10 👍 7
Another Blueprint In The Wall: How to Ask Frontier AI Like a Kid?
This paper reports experiments across six frontier model types from OpenAI, Anthropic, xAI, and Google DeepMind. Ten independent sessions per model type used th…
更多論文（arXiv / ChatPaper / Semantic Scholar，30 篇）
- 1 Semantic Scholar
Analysing Structures of Silence in AI-mediated Public Spheres
Analysing Structures of Silence in AI-mediated Public Spheres. Philosophy & Technology…
- 2 Semantic Scholar
Psychological barriers and pedagogical implications of artificial intelligence-induced anxiety among English lecturers in higher education
Psychological barriers and pedagogical implications of artificial intelligence-induced anxiety among English lecturers in higher education. Discover Education…
- 3 Semantic Scholar
Evolving digital skill demands before and after generative AI: a global network analysis via upwork data
Evolving digital skill demands before and after generative AI: a global network analysis via upwork data. Data mining and knowledge discovery…
- 4 Semantic Scholar
Early evidence for a multi-agent AI simulator for clinical reasoning practice: performance, consistency, and challenges
Early evidence for a multi-agent AI simulator for clinical reasoning practice: performance, consistency, and challenges. Diagnosis…
- 5 Semantic Scholar
Adaptive versus non-adaptive sampling for Gaussian-RBF surrogates: a replicated benchmark across analytical and differential-equation models
Adaptive sampling extends a nested design point by point, but repeated surrogate fitting and acquisition search are justified only when they improve prediction …
- 6 Semantic Scholar
Agentic AI in educational environments and its association with social and motivational pathways to learning performance
Agentic AI in educational environments and its association with social and motivational pathways to learning performance. Discover Education…
- 7 Semantic Scholar
Human-AI Interaction: Evaluating LLM Reasoning on Digital Logic Circuits and the Perception-Correctness Gap
Human-AI Interaction: Evaluating LLM Reasoning on Digital Logic Circuits and the Perception-Correctness Gap. SN Computer Science…
- 8 Semantic Scholar
SE4SC-LLM: an LLM-Augmented symbolic execution framework for smart contracts
SE4SC-LLM: an LLM-Augmented symbolic execution framework for smart contracts. International Conference on Automated Software Engineering…
- 9 Semantic Scholar
The SkySentience conceptual framework for accountable drone decision support in public safety
The SkySentience conceptual framework for accountable drone decision support in public safety. Discover Informatics…
- 10 Semantic Scholar
A multi-site benchmarking framework for scalable extraction of geriatric care constructs from electronic health records
A multi-site benchmarking framework for scalable extraction of geriatric care constructs from electronic health records. npj Health Systems…
- 11 Semantic Scholar
Genolator enables protein function interpretation using a multimodal large language model fusing genomic and structural interpretation with natural language interaction
Genolator enables protein function interpretation using a multimodal large language model fusing genomic and structural interpretation with natural language int…
- 12 Semantic Scholar
A system dynamics approach to sustainable monetization in digital platforms: balancing advertising and user retention in mobile gaming
A system dynamics approach to sustainable monetization in digital platforms: balancing advertising and user retention in mobile gaming. Future Business Journal…
- 13 Semantic Scholar
Large Language Models im Umfeld der maschinellen Übersetzung. Eine Analyse des Potenzials von ChatGPT-4o zum Pre-Editing in der Wirtschaftsübersetzung
Large Language Models im Umfeld der maschinellen Übersetzung. Eine Analyse des Potenzials von ChatGPT-4o zum Pre-Editing in der Wirtschaftsübersetzung.…
- 14 arXiv
Agentic Societies Need a Social Harness
An agentic society is a collection of AI agents that coordinate autonomously across trust boundaries, on behalf of different principals whose objectives may onl…
- 15 arXiv
When Should LLMs Abstain? Chain-of-Self-Questioning for Selective Risk Control
Large language models can produce fluent answers when their factual support is weak. This paper introduces Chain-of-Self-Questioning (CoSQ), a prompt-only frame…
- 16 arXiv
What Breaks Under Pruning in Smart Homes, and When? Evaluating LLM Degradation Across Architectures and Task Complexity
Pruning can reduce the deployment cost of large language models (LLMs), but its impact on context-grounded tool calling remains poorly understood. We systematic…
- 17 arXiv
LACE: Layer-Wise Compression for Dynamic Frame Rate Codecs
Neural audio codecs are a key component in speech language modeling. However, their high frame rates lead to long sequence lengths, increasing computational cos…
- 18 arXiv
ENCP: Episode-Normalized Conformal Prediction for Vision-and-Language Navigation
Uncertainty estimation for Vision-Language-Navigation (VLN) models is a critical task since it can help identify ambiguous and unreliable predictions, enabling …
- 19 arXiv
Verifiable Social Reasoning for LLM Assistants
LLM assistants are widely used for daily social advice, yet evaluating their social reasoning in such consultation settings remains challenging since (i) it req…
- 20 arXiv
FreqSpaNet: Frequency and Spatial Learning of SFPF for Physical Layer Hardware Integrity Detection
Unauthorized hardware replacement can preserve a wireless device's logical identity while altering its physical implementation, posing a challenge to hardware i…
- 21 arXiv
LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence
We introduce LimiX-2, a new model in the LimiX family, developed through model and data scaling guided by our previously established scaling laws. LimiX-2 adopt…
- 22 arXiv
Bridging the Gap Between Homogeneous and Heterogeneous Asynchronous Optimization Is Surprisingly Difficult
Modern large-scale machine learning tasks often require multiple workers, devices, CPUs, or GPUs to compute stochastic gradients in parallel and asynchronously …
- 23 arXiv
Det-LIME: Detector-Aware, Multi-Instance Local Interpretable Model-Agnostic Explanations for Automated Marine Mammal Detection
Despite the rapid uptake of black-box object detectors in marine mammal research and monitoring, explainability techniques are rarely integrated into conservati…
- 24 arXiv
Bias-Induced Crossover in Absolute Capacity of Dense Associative Memory
The absolute capacity of dense associative memory has mainly been analyzed for unbiased patterns. Here we examine the effect of bias in centered binary patterns…
- 25 arXiv
JustFit: 200K-Token LLM Serving on a 24 GiB Laptop with Just-in-Time State Management
Capable open-weight models make local coding and reasoning attractive, but their context and execution state strain laptop memory. We present JustFit, an MLX-ba…
- 26 arXiv
Coupled Calibration and Learning: Mitigating Teacher Bias in LLM Distillation without Target-Domain Reward Feedback
Large language model (LLM) distillation aims to transfer the capabilities of a powerful teacher to a smaller student. Direct imitation, however, can also transf…
- 27 arXiv
Decomposition Buys Integrity, Not Yield
Multi-agent systems split a task across a tree of agents and justify the split with folklore: smaller contexts, cleaner separation, parallelism. We ask what the…
- 28 arXiv
Tables Decoded: DELTA for Structure, TARQA for Understanding
Table understanding is a core task in document intelligence, encompassing two key subtasks: table reconstruction and table visual question answering (TabVQA). W…
- 29 arXiv
ORCA: Occlusion-Aware Refinement and Completion for Novel View Synthesis
Novel-view synthesis from a single image is a fundamentally ambiguous problem. As the camera moves away from the input viewpoint, previously hidden regions beco…
- 30 arXiv
BrainFocus: EEG-Guided ROI Selection for Efficient Vision-Language Models
Vision-language models (VLMs) achieve strong visual question answering (VQA) performance, but processing large cluttered images is computationally expensive whe…