Papers

New on arXiv

Wed 19 Aug 2026

356 papers

Security and Privacy in Large-Model-Driven Embodied Agents: Attacks, Defenses, and Future DirectionsLLM agents and tool use → Agent threats, guardrails and governance
An Empirical Cost Attribution of Context-Compression Gateways in Multi-Turn Coding AgentsLLM agents and tool use → Agent memory systems +1
EMODY Flow: Emotion-Aware Audio-Driven Full-Body Motion GenerationDiffusion models → Audio, speech and music
MonoMoE: An Efficient Fused Mega-kernel for Quantized MoE DecodingMixture-of-experts → MoE large language models
A Generalized Optimization Engine (GOE) for Edge AI Inference AccelerationEfficient LLM inference and serving → Serving systems and request scheduling
Squeezing the Cache, Preserving the Truth: Monotonic Equipotential Allocation with Geodesia-KVEfficient LLM inference and serving → KV cache optimization
ViTexSZ: Heterogeneous Vision-Text Knowledge Distillation for EEG Seizure DetectionKnowledge distillation → Distillation for medical imaging
Agentic AI for Safety-critical Multi-drone Systems: Challenges and OpportunitiesLLM agents and tool use → Multimodal and embodied agents
Who Delegates to AI? Evidence from Agent Configurations in GithubLLM agents and tool use → Software engineering coding agents
An Agentic RAG and Evaluation Framework for Assurance Case Generation: Industrial Use Case for the EU Cyber Resilience Act ComplianceLLM agents and tool use → Agent threats, guardrails and governance
A Plug-in Interpretation of Conditioning in Score-Based Diffusion ModelsDiffusion models → Diffusion foundations and sampling
Beyond Multimodal Alignment: Certifying Physical Language through Response Substitution and Ordered ExecutionWorld models → World models for language agents
Fine-Tuning VLAs with Self-Demonstrated Generative Control for Multi-Task ManipulationVision-language-action models → Generalist VLA models and benchmarks
Accelerated Genetic Programming Hyper-Heuristics for Simulation-Based Scheduling via Agentic AILLM agents and tool use → Software engineering coding agents
Beyond Imitation: Filtering On-Policy Distillation by Reasoning ProgressLLM reasoning → Prompting and chain-of-thought reasoning +1
HYDRA: A Heterogeneous Chiplet DSE Framework for Serving Dynamic Hybrid LLM WorkloadsEfficient LLM inference and serving → Serving systems and request scheduling
Concentrated Liquidity Provision: a Reinforcement Learning PerspectiveDeep reinforcement learning methods → Adversarial robustness and RL in finance
CAViAR: A Causal Video Dataset for Fine-Grained Accident Reasoning in Real-World ScenariosLLM reasoning → Chain-of-thought reasoning for driving
Does Marginal Coverage Guarantee Class-Conditional Safety for Zero-Shot VLMs Under Shift?Multimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Learning the Right Abstraction: Neural Reduced Dynamics for Complex Robot ControlDeep reinforcement learning methods → RL for robot locomotion and manipulation
Linguistic Holonomy and Statistical Watermarks: Inner Geometry of Meaning-Preserving TransformationsLLM safety and security → Watermarking generated content and models
GRACE: Grounded Reasoning via Adapter Composition and Evidence-Aware Calibration for Educational Visual Question AnsweringLLM reasoning → Multimodal reasoning
Mechanistic Tomography: Designed Measurement for Control-Oriented InterpretabilityMechanistic interpretability → Circuits and mechanistic analysis
Clustering and Token Denoising for Faster and More Robust VLMsMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
SPADE: Self-Play in Adaptive Synthetic Executable EnvironmentsSynthetic data and self-improvement for LLMs → Self-evolving LLM agents
ADEPT: Accelerating Dexterity via Pre-Training and Post-Training using Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
Beyond Teacher Likelihood: Group-Calibrated On-Policy Distillation for Long-Context ReasoningKnowledge distillation → On-policy and self-distillation for LLMs
Beyond the Transcript: Detecting Covert Co ordination in Latent Multi-Agent CommunicationLLM agents and tool use → Agent threats, guardrails and governance
Continuous-Time Reinforcement Learning for Controlled Hawkes Jump-DiffusionsDeep reinforcement learning methods → RL theory and sample complexity
Pre-Compiled Pipeline Shards for Distributed LLM Inference on Intel AI PC FleetsEfficient LLM inference and serving → Serving systems and request scheduling
PGFS++: Molecular Property Improvement under Synthesis and Diversity ConstraintsDeep reinforcement learning methods → RL for energy, networks and operations
Discretizing Continuous Time Series for Imputation with Masked Diffusion TrainingDiffusion models → Flow matching, time series and science
Open-MOPD: Diagnosing and Fixing Capability Imbalance in Multi-Teacher On-Policy DistillationKnowledge distillation → On-policy and self-distillation for LLMs +1
DA-WAM: Decision-Aligned Future Latents for Driving World ModelsWorld models → World models for autonomous driving +1
ReWEIGH the Evidence: Calibrating Token-Level Ordinal Visual Evidence to Mitigate Hallucinations in Large Vision-Language ModelsMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Diffusion Models for High-Dimensional Clustered Data: Intrinsic-Dimension Adaptivity via Bayesian ClassificationDiffusion models → Diffusion foundations and sampling
GS-VLA: Plug-and-Play Viewpoint Canonicalization for Frozen VLA Policies via Gaussian SplattingVision-language-action models → 3D and spatial-aware VLA
Grading the Graders: Verification Autonomy Levels (L0-L5) for LLM ReasoningLLM reasoning → Reasoning benchmarks and evaluation
A Theory of Post-hoc Debate JudgementLLM agents and tool use → Social simulation and multi-agent debate
Mise-en-Sc`ene: Implicit Layout Emergence in Diffusion Transformers for Human-AI Design Co-CreationDiffusion models → Image generation, text-to-image and editing
Uncertainty-Aware Art-Historical Dating with Vision-Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
rEDMRec: Distilling Large Language Model Reasoning into an Editable Experience Memory for RecommendationLLM reasoning → LLM reasoning for recommendation +1
Breaking the weakest link to evade vision language modelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
MedUAG: Unified Understanding and Generation for Medical Multimodal ModelsMultimodal LLMs / vision-language models → Medical vision-language models
SkillForge: Self-Distilling Agents for Project-Specific Issue ResolutionLLM agents and tool use → Software engineering coding agents
Test-Time Scaling in the Wild: Why Exploitation, Not Exploration, Is the BottleneckLLM reasoning → Test-time compute and efficient thinking
SMTrap: Cost-Effective DoS Attacks Against Large Reasoning Models via SMT Conflict GuidanceLLM reasoning → Reasoning for agents, planning and code
Training-Free Inference-Time Self-Reflection and Cost-Bounded Early Stopping for Large Language ModelsLLM reasoning → Reinforcement learning for reasoning
Falcon Perception-HD: High Density Perception via Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
DentAgent: Evidence-Centric Multi-Agent Coordination for Multimodal Dental ReasoningLLM reasoning → Clinical and biomedical applications +1
Think-to-Personalize: Unifying Reasoning and Retrieval for User-Centric Personalized Dense RetrievalLLM reasoning → Knowledge- and tool-augmented reasoning
ORBITER: Conflict-Aware Decision-Making for Agentic Last-Mile DeliveryLLM agents and tool use → Social simulation and multi-agent debate
EVADE: Evidence-Verified Agentic Diagnosis with EscapeLLM reasoning → Clinical and biomedical applications +2
To Go Far, Go Together: Diverse Preferences Induce a Curriculum for Reward OptimizationLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Decision-Metric Alignment in Latent World Models: Diagnostics and Action-Conditioned Objectives for MPC PlanningWorld models → Latent predictive world models
A Multi-Agent Platform for Automated Enterprise Analytics and Insight GenerationLLM agents and tool use → Web, search, and research agents
CL4D: Contrastive Language-4D Pretraining for Vision-Language Reasoning in Dynamic ScenesLLM reasoning → Spatial and embodied reasoning
DocClaw: A Unified Agentic System for Intelligent Document ProcessingLLM agents and tool use → Web, search, and research agents
RTPO: Reverse-Turn Policy Optimization for Stabilizing Agentic RL TrainingLLM reasoning → Reinforcement learning for reasoning +1
Learning What to Fail On: Failure-Mode Contextual Bandits for Adversarial Data CurationRetrieval-augmented generation → Privacy and security of RAG
Sanyu Studio: A Multi-Agent System for Art-Historical Narrative ConstructionLLM agents and tool use → Multimodal and embodied agents
DynCur-Geo: Dynamic Curiosity Reward Shaping for Multimodal Active Geo-LocalizationDeep reinforcement learning methods → RL for robot navigation and UAV control
Vision-Language Models for Egocentric Video: From Hand-Object Interaction to Embodied AIMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Reinforced Planning with Latent World ModelsWorld models → Latent predictive world models
Progressive Experience Fusion for Multi-Task World Model Control in Endovascular NavigationWorld models → Medical and surgical world models
Code Health in LLM-Based Test Generation: Effectiveness and Token EfficiencyLLM agents and tool use → Software engineering coding agents
Report on The 1st Workshop on Human-Centered Proactive and Personalized Agents for Interactive Information Access at CHIIR 2026LLM agents and tool use → General LLM agent frameworks and benchmarks
PILOT Technical ReportLLM agents and tool use → Agentic recommender systems
Preference Reasoning under Indeterminacy in Large Language ModelsLLM reasoning → Reasoning benchmarks and evaluation +1
When Safety Overrides Vision: Exploring Dynamics between Vision Influence and Safety Alignment in Vision-Language ModelsLLM reasoning → Multimodal reasoning +3
PCQA-R1: Advancing Generalized 3D Point Cloud Quality Assessment with Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
CTIFoundry: An Agent-Native Corpus Scaffold for Cyber Threat IntelligenceLLM agents and tool use → Agent threats, guardrails and governance
VA-Judger: Reward Modeling from Human Preference Feedback for Joint Video-Audio GenerationLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
Infrared Universality of Collective Dynamics across Transformer and State-Space ArchitecturesState-space sequence models (S4, Mamba) → Core SSM architectures for language
Can a Lightweight Multimodal Model Estimate LLM Reasoning Performance? A Study for Compute-Optimal Document InferenceLLM reasoning → Multimodal reasoning
FD-CanKD: Frequency-Decoupled Cross-Attention Distillation as a Refinement Prior for Compact Object DetectorsKnowledge distillation → Detection and dense prediction distillation
OmniHandwritingOCR: A Diagnostic Benchmark for Evaluating Multimodal LLMs in Handwritten OCR ScenariosMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
From Storage to Access: Verifiable Activation of Parametric Knowledge in LLMs via Explicit Priming and Implicit ReasoningLLM reasoning → Reasoning benchmarks and evaluation
MR-IQA-2: Faithful Image Quality Reflection via Fine-Grained Credit AssignmentMultimodal LLMs / vision-language models → Image editing and quality assessment
Beyond LLM-Based Reasoning: Lightweight GNNs for Agent Failure AttributionLLM reasoning → Reasoning for agents, planning and code +1
Beyond Forgetting: Diagnosing and Harnessing Shared Reasoning in Continual RLVRLLM reasoning → Reinforcement learning for reasoning
SemaPLC: A Project-Grounded, Verification-Gated Agent Harness for PLC Code GenerationLLM agents and tool use → Software engineering coding agents
Bridging Search and CRM: Productionizing AI Product Research Agents for Customer Re-EngagementLLM agents and tool use → Social simulation and multi-agent debate
FinRCA-Bench: Benchmarking Evidence Retrieval and Reasoning for Financial AI SystemsLLM reasoning → Reasoning for agents, planning and code
StateTrace: An Object-Centric Framework for Hidden-State Spatiotemporal Reasoning in Long VideosLLM reasoning → Multimodal reasoning
DART-SD: Diamond-topology Aware Retrieval and Tuning for Self-Distillation of Multi-Turn Tool-Calling AgentsLLM agents and tool use → Tool learning and function calling +1
Science Done on a Machine by a Machine: AI Agents in Computational ChemistryLLM agents and tool use → Medical and scientific domain agents
UMER: Unifying Embedding and Ranking via Pair-Aware Discriminative Reasoning for Universal Multimodal RetrievalLLM reasoning → Multimodal reasoning
Bayesian Partner Modelling enables Adaptive Replanning for LLM CoordinationLLM agents and tool use → Agent skill learning and evolution
MissDiag: Diagnostic Evaluation of Incomplete-Knowledge Robustness in KGQA and KG-RAGRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
Partition the Support, Reconstruct the Residual: Training-Free Sparse Attention for Video Generation and World ModelsWorld models → Video-generation world models and simulators
A Locally Deployable Tool-Grounded LLM Multi-agent Framework for Automating Methane Emission Analysis and ReportingLLM agents and tool use → Web, search, and research agents
Optimization of the Railcar Assignment Problem Using Zone-based Double Deep Reinforcement LearningDeep reinforcement learning methods → RL for autonomous driving and traffic
The Embodiment Gap in Robot Foundation ModelsVision-language-action models → Generalist VLA models and benchmarks
FM-Bench: A Benchmark for Long-Horizon Management with Competing AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Mechanistic Interpretability of Structure-Aware Numerical Reasoning in LLaMA 3.1 8BLLM reasoning → Reasoning benchmarks and evaluation +1
Role-Conditioned Sub-Token Routing for Efficient Vision-Language-Action PoliciesVision-language-action models → Efficient VLA inference and adaptation +1
Vector Symbolic Policy GradientDeep reinforcement learning methods → Deep RL algorithms and representations
LEDGER: Claim-to-Evidence Trace Graphs for Auditing LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance
Theoretical Study on the Evidential Learning-based Variational Autoencoder

All 356