Papers

New on arXiv

Sun 13 Sep 2026

241 papers

Rethinking Streaming Video Diffusion Model: Context, Execution, and TrainingDiffusion models → Video generation
The Imitation Game: When LLMs Learn to Reason Like Programs via Code-Centric Reasoning Data SynthesisLLM reasoning → Reasoning for agents, planning and code +1
World-Action Models for Robot Learning and Control: A SurveyWorld models → Embodied and robot world models +1
The Immutable Past: Formalizing State Mutability and Conflict Resolution in Mutable RAGRetrieval-augmented generation → Agentic RAG, agents and memory
World Models for Cross-Machine CNC Transfer under Partial Sensor OverlapWorld models → Embodied and robot world models
You Don't Need To Train: Agentic Heuristic Learning Studio for Executable Human Activity RecognitionLLM agents and tool use → Multimodal and embodied agents
Towards Scalable RLVR: Multimodal Instruction Following Data Synthesis and DistillationSynthetic data and self-improvement for LLMs → Self-play RL for reasoning
OmniHarness: Harnessing Generalizable Visual Generation via Symbolic Policy LearningMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
State of Thought Enables Endogenous ReasoningLLM reasoning → Test-time compute and efficient thinking
Retrieval-Driven Memory Reconsolidation for Long-Term LLM AgentsLLM agents and tool use → Agent memory systems
Self-Orchestrating Language Models: Leveraging Semantic Dependence for Efficient InferenceDiffusion models → Discrete and text diffusion
El Agente Potente: High-Throughput Agentic Atomistic SimulationsLLM agents and tool use → Medical and scientific domain agents
One Model, Two Physical Stories: Auditing Misalignment in Multi-Modal World ModelingWorld models → Latent predictive world models
ANASSA: An Agentic AI Orchestration Framework for Spatial IntelligenceLLM agents and tool use → Multimodal and embodied agents
Zero-shot video highlight detection based on text descriptions and synthetic imagesDiffusion models → Video generation
The Stochastic Deputy: Structural Tenant Isolation for Tool-Using LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance
Func-R1: Incentivizing Mathematical Function Reasoning in Multimodal Large Language ModelsLLM reasoning → Multimodal reasoning +1
Loop-Back Authority in LLM Agent Teams: A Paired Experiment on Flat and Hierarchical CoordinationLLM agents and tool use → General LLM agent frameworks and benchmarks
TriCalRAG: A Three-Strategy, Retrieval-Augmented Benchmark for On-Premise LLM-Based Root Cause Analysis in AIOpsRetrieval-augmented generation → Privacy and security of RAG
Fabrication After Tool Failure: Tool-Augmented Agents Assert Values Their Tools Did Not ReturnLLM agents and tool use → Tool learning and function calling
Building Legal Reward Models for Grounding and AbstentionRetrieval-augmented generation → Legal and financial RAG +1
CALICO: A Human-Centered, Codebook-Aligned System for AnnotationLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
CrossDistill: Balancing Quality and Diversity via Trajectory-Level Hybrid Few-Step DistillationDiffusion models → Diffusion foundations and sampling
Detecting and Localizing Segment-Level Poisoning in Multi-Source LLM-Agent InputsLLM agents and tool use → Agent threats, guardrails and governance
A Two-Dimensional Study of the Model Context Protocol: Publication and AdoptionLLM agents and tool use → Agent threats, guardrails and governance
Carryover Drafting: Recycling Rejected States for Speculative DecodingEfficient LLM inference and serving → Speculative decoding
An immune world model for multiscale forecasting and therapeutic hypothesis generationWorld models → Medical and surgical world models
Lightning Weave: Improving the Accuracy-Efficiency Frontier of Reasoning Models through Capability CompositionLLM reasoning → Test-time compute and efficient thinking
Perceive, Refine, Reason: A Calibrated Pipeline for Measuring Indicators in Strategic Visual Communication on Social MediaLLM reasoning → Multimodal reasoning
ViTeGate: Visual-Textual Triggered Knowledge Poisoning for Vision-Language Retrieval-Augmented GenerationRetrieval-augmented generation → Privacy and security of RAG
Compositional SVG Generation via VLM-Driven Hierarchical Semantic ParsingLLM reasoning → Multimodal reasoning +1
CIG-MIA: Context-Induced Information Gain Membership Inference Attacks against Retrieval-Augmented GenerationRetrieval-augmented generation → Privacy and security of RAG +1
Optimizing Sparse Outcomes Through Dense Behavioral Signals via Value-Guided Preference DistillationLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
BigMoMo: Efficient Inference of Large-Scale MoE with Speculative Decoding on Mobile DevicesMixture-of-experts → Efficient MoE training and inference systems +1
Diffusion-Based Generation of Gait TrajectoriesDiffusion models → Video generation
DynSTEER: Dynamic Stage-wise Trajectory Evaluation and Execution-time Review for AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Know When to Stop, Where to Restart: Accelerating Multi-Turn Agentic On-Policy DistillationLLM agents and tool use → Reinforcement learning for LLM agents
LLM Agent Capabilities Should Follow Task Intent and Context SourceLLM agents and tool use → Agent threats, guardrails and governance
Beyond Benchmark Scores: How Synthetic and Authentic Query Distributions Diverge in RAG EvaluationRetrieval-augmented generation → RAG pipeline design and evaluation
AlgoRAG: Retrieval-Augmented Generation for Theoretical Computer Science Education -- A Comprehensive Evaluation Framework for Algorithm Analysis and Complexity TheoryRetrieval-augmented generation → Agentic RAG, agents and memory
Disentangling Topology and Diversity in Multi-Agent LLMs for Multilingual Low-Resource Emotion DetectionLLM agents and tool use → Social simulation and multi-agent debate
GLAM: Training a latent world model over global spatiotemporal memory for active exploration and navigationWorld models → World models for navigation
NavPatch: Evidence-Guided Object-Level Costmap Correction with Vision-Language ModelsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Theseus in the Graph: Towards Traceable Multi-Hop Graph NavigationLLM reasoning → Knowledge- and tool-augmented reasoning
Selective Tool Use for Agentic Change Visual Question Answering in Remote SensingLLM agents and tool use → Multimodal and embodied agents
Physically Partitioned KVCache Format for CPU--GPU Load Balancing in MoE InferenceMixture-of-experts → Efficient MoE training and inference systems +1
AlayaVista: Streaming World Modeling from Panoramic States to Perspective VideoWorld models → Video-generation world models and simulators
A latent dimension of Condorcet's jury theorem for multiple AI advisersLLM reasoning → Test-time compute and efficient thinking
Differentiable Digital Signal Processing Mixture Model-Guided Diffusion for Synthesis Parameter Estimation from Harmonic Sound MixturesDiffusion models → Audio, speech and music
Safety Signals to Verify NetOps Agents with Action-Level GranularityLLM agents and tool use → Software engineering coding agents
A Hybrid Dependency-Aware Framework for Task Decomposition and Dynamic Agent Generation in Oracle-to-PostgreSQL MigrationLLM agents and tool use → Web, search, and research agents
Question's Gambit: The First Move Matters in Agentic Deep SearchLLM agents and tool use → Web, search, and research agents
Dynamic Learning Solutions: A System for Personalized Educational Video GenerationRetrieval-augmented generation → Multimodal RAG
MOSCOPT: Mixture-of-Skills Collective Optimization for LLM AgentsLLM agents and tool use → Agent skill learning and evolution
Two-Stage Mixture-of-LoRA for Multi-Task Medical Vision-Language LearningMultimodal LLMs / vision-language models → Medical vision-language models
Flattening Every Memory Peak in Long-Context Mixture-of-Experts TrainingMixture-of-experts → Efficient MoE training and inference systems
E2A-Bench: Benchmarking Evidence-to-Action Reliability in Financial Chart ReasoningLLM reasoning → Reasoning for agents, planning and code
AnnoSketch: Evaluating and Collecting Human Sketches for MLLM-assisted Chart AnnotationMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
Vision-Language Models for Criterion-Level Grading of Handwritten Examinations in Outcome-Based EducationMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
VGFM: Expressive Robot Policies via Dense Value Guidance in Flow MatchingDiffusion models → Diffusion and flow policies for robotics +1
SPARK: Representation-Level KV Memory Alignment for Safer Vision-Language ModelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
DenMark: Robust Semantic Watermarking for Diffusion Language ModelsDiffusion models → Robustness, privacy and adversarial aspects +1
The Attribution-Compression Frontier in Retrieval-Augmented GenerationRetrieval-augmented generation → RAG robustness and adaptive retrieval
CoArena: Evaluating Computer-Use and Multi-Agent Systems in Real TimeLLM agents and tool use → General LLM agent frameworks and benchmarks
OpWeave: Flexible Operator Disaggregation for Heterogeneous LLM ServingMixture-of-experts → Efficient MoE training and inference systems +1
Enhancing Human Mobility Prediction with Spatially Aware LLM-based Multi-Agent SystemsLLM reasoning → Spatial and embodied reasoning +1
ECAS: An Edge-Controlled Agentic System for Validation-Gated Scientific Application ExecutionLLM agents and tool use → Software engineering coding agents
Learning to Refer from Estimated Listener GazeMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
From Hand-Crafted to LLM-Based Variation Operators in Metaheuristics: A Tutorial
Energy Vision--Language--Action: A Controlled Multimodal Benchmark for Intent-Conditioned Residential Energy Management
SignTrace: Describe a Sign, Find the Word
SlideLab: Audience-Centered Scientific Slide Generation and Evaluation
Cartograph: Federated Tool Discovery with Operator-Attested Retrieval for AI Agents
ORDER: A Fictitious-World Benchmark for Domain-Adaptive Embodied AI
Complementary rPPG-Derived and Lip-Region Frequency Cues for Talking-Face Deepfake Detection
Brain-to-Image Generation: Reconstructing Visual Stimuli from EEG using Generative Adversarial Networks
Performance vs Consistency: Evaluating a Foundation Model in Lung-RADS Screening
Making Political Text Scaling Comparable: Infrastructure and Hyperparameter Sensitivity for 17 Algorithms
Decentralized Optimal Equilibrium Learning Over Dynamic Networks
Structure is not mechanism: high-gain gated-FFN rows across text and genomic foundation models
Pseudo-Label Augmentation for Affect Sensing in Small Collaborative Groups
AssemblyGrid v1: A Benchmark for Multi-Robot Production with Temporary Coalitions, Local Information, and Geometric Constraints
Efficient Multimodal Generative Recommendation with Latent Narrative Reasoning
Beyond Distribution Matching: Semantics-Consistent Tabular Diffusion with Weak Semantic Priors
A Dynamic Aggregation Strategy Enhanced Efficient Global Optimization Algorithm for Solving High-Dimensional Turbomachinery Design Problems
A panoramic aerodynamic performance prediction method for turbomachinery cascades using transformer-enhanced neural operator
Signed p-adic Residual Encodings of Finite-Domain All-Different Systems with a Sudoku Case Study
Digital Persuasion: Understanding the Impact of Online Influencers on Public Opinion
POSPAN: Position-Constrained Span Masking for Language Model Pre-training
HintMiner: Automatic Question Hints Mining From Q&A Web Posts with Language Model via Self-Supervised Learning
Driver Behavior Estimation at Signalized Intersections Using a Physics-Constrained Decision-Conditioned Autoregressive Transformer
Managing Action Preconditions in Neuro-Symbolic RL: Three Placement Strategies for Embodied Agents
Causal neural set filtering for online multi-target tracking
"Looking for Something Weird to Happen": How Humans Sustain AI Agent Novelty Amid Semantic Collapse
LLMs as Oracles: Reliance on LLMs for Subjective Personal Questions
Prescreening Point Defects in Semiconductors With Machine Learning
A Responsive Present, a Shared Past, a Social Other: Teens' Overreliance on Companion AI Chatbots
Tackling Failure Modes of PINNs and PIKANs Using Conflict-Free Gradients
Efficiency Hallucination: Formalizing and Measuring Behavioral Calibration in LLM-Based Code Optimization
Enemray: Toward Capable Language Models for Hassaniya

All 241