Papers

New on arXiv

Sun 5 Jul 2026

211 papers

When Retrieval Fails Before It Begins: Structurally Indirect Prerequisite Eviction as a Retention Failure in Agentic MemoryLLM agents and tool use → Agent memory systems
SiGMA: Sign-Guided Merging and Adaptation for Multimodal Continual Instruction TuningMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
HRIBench: Benchmarking Interaction-Centric Human-Robot CollaborationVision-language-action models → Failure recovery and robustness
LLM-Centric Agentic AI for UAV Swarms: Architecture, Enabling Technologies, and Open ProblemsLLM agents and tool use → Multimodal and embodied agents
Learning to Control LLM Agent Harnesses with Offline Reinforcement LearningLLM agents and tool use → Software engineering coding agents +1
Prompt-to-Paper: Agentic AI System for BioinformaticsLLM agents and tool use → Web, search, and research agents
Mask2Real-WM: Segmentation Masks as a Sim-to-Real Bridge for Controllable Dexterous World ModelsWorld models → Embodied and robot world models
Auto: The AGI CompilerLLM agents and tool use → Software engineering coding agents
CRISP: A Spatiotemporal Camera-Radar Backbone for Driving via Forecasting-Based World-Model PretrainingWorld models → World models for autonomous driving
Measuring Harness-Induced Belief Divergence in Multi-Step LLM AgentsLLM agents and tool use → Software engineering coding agents
VLA Grounder: Language-Conditioning Space Optimization for Black-Box VLA ModelsVision-language-action models → Reinforcement learning and post-training
Constrained Flow Matching via Lagrangian Dual FlowsDiffusion models → Flow matching, time series and science
Compressing the Validation Bottleneck: An Agentic Self-Driving Lab for Scientific DiscoveryLLM agents and tool use → Web, search, and research agents
Why Pure Reasoning is Not Enough: Nature as the Source of Mathematical InnovationLLM reasoning → Mathematical reasoning
Geographic Diversity Beats Data Volume for Cross-Domain Generalization in Zero-Label JEPA Driving World ModelsWorld models → World models for autonomous driving +1
UniSkip-Mamba: A Frequency-Aware State Space Model for Audio-Visual Temporal Forgery LocalizationState-space sequence models (S4, Mamba) → Speech and audio SSMs
TrustCLIP: Learning Private Visual Features via Adversarial ReconstructionMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
Regime-Conditional Stabilisation of LLM-Augmented Cooperative Multi-Agent Reinforcement LearningLLM agents and tool use → Reinforcement learning for LLM agents
Don't Commit Alone: Joint Token Commitment in Diffusion Language ModelsDiffusion models → Discrete and text diffusion
Operator-on-F complements value-equivalence: a planning-time diagnostic for latent world modelsWorld models → Latent dynamics models and model-based RL +1
Flash-BoN: Instant Drafts for Inference-Time Scaling in Diffusion ModelsLLM reasoning → Diffusion reasoning and test-time scaling +1
Tightening the Score Matching Gap for Diffusion ModelsDiffusion models → Diffusion foundations and sampling
A Retrieval-Augmented Framework for Detecting and Resolving Pragmatic Ambiguities in Natural Language RequirementsRetrieval-augmented generation → RAG pipeline design and evaluation
Autonomous Information Seeking: A Roadmap for Agentic Recommender SystemsLLM agents and tool use → Agentic recommender systems
dOPSD: On-Policy Self-Distillation for Diffusion Language ModelsDiffusion models → Discrete and text diffusion
ACE-Brain-0.5: A Unified Embodied Foundational Model for Physical Agentic AILLM agents and tool use → Multimodal and embodied agents
UI-MOPD: Multi-Platform On-Policy Distillation for Unified GUI AgentsLLM agents and tool use → GUI and mobile agents +1
Learning Task-Sufficient World Models by Synergizing Agentic Exploration and Structured ModelingWorld models → World models for language agents +1
NKI-Agent: Domain-Specific Fine-Tuning and Agentic Tool Use for Neuron Kernel GenerationLLM agents and tool use → Software engineering coding agents
MechMath Agent Team: LLM Driven Agents for Mathematical ResearchLLM agents and tool use → Educational agents and math problem solving
Memory-Orchestrated Semantic System (MOSS): An Auditable Agentic Memory ArchitectureLLM agents and tool use → Agent memory systems
Knowledge Base Poisoning Attacks and Defense for Policy-Aware LLM-RAG FrameworkRetrieval-augmented generation → Privacy and security of RAG
Nemotron-Labs-3-Puzzle-75B-A9B: Compressing Hybrid MoE LLMsMixture-of-experts → Efficient MoE training and inference systems
RL Forgets! Towards Continual Policy OptimizationMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP +1
Optimal Mixture-of-Experts Model Averaging for Conditional Generative ModelsMixture-of-experts → MoE theory and gating functions
WPG-MoE: Weak-Prior-Guided Dense Mixture-of-Experts for User-Level Social Media Depression DetectionMixture-of-experts → Multimodal MoE models
IRIS: An Intelligent Vision-Language System for Ocular Surface Diseases via Topic Tree and Scene-Driven VQA GenerationMultimodal LLMs / vision-language models → Medical vision-language models
One Framework for All: Cross-Modal Membership Inference for Generative ModelsLLM safety and security → Membership inference attacks
Do GUI Agents Believe Their Eyes? Diagnosing State-Belief Reliance on Pixels versus StructureLLM agents and tool use → GUI and mobile agents
Agent-driven Long-tail Simulation for Autonomous DrivingEnd-to-end autonomous driving → Large language models for driving +1
HAS-Bench: Evaluating LLM-Based Human-Agent Systems under Configurable Human ParticipationLLM agents and tool use → General LLM agent frameworks and benchmarks
SAD-LoRA: Spectral Alignment for Low-Rank Knowledge DistillationKnowledge distillation → General knowledge distillation methods
CausalGame: Benchmarking Causal Thinking of LLM Agents in GamesLLM reasoning → Causal reasoning in LLMs +1
Agentic SABRE: An Uncertainty-Aware Neuro-Symbolic Multi-Agent Framework for Adaptive Ransomware DetectionLLM agents and tool use → Agent threats, guardrails and governance
Agentic-V2X: Small Language Model Agents for Deadline-Aware V2X Scheduling in 5G/6G NetworksLLM agents and tool use → Agentic AI for wireless networks
Risk-Constrained Freshness-Aware Semantic Caching for Open-Web Retrieval-Augmented LLMsRetrieval-augmented generation → Generation and context integration
Self-Reference in Large Language Models: The Introspection Threshold for Recursive Self-ImprovementSynthetic data and self-improvement for LLMs → Self-rewarding alignment and self-training
EMPURPLE: A Free Lunch for Diffusion Distillation based on the Information BottleneckDiffusion models → Diffusion foundations and sampling
HALO-WA: Hybrid-Attention Latent-Guided Online Reinforcement Learning for World-Action ModelsWorld models → Embodied and robot world models +1
Quantize the Target, Quantize the Drafter: Efficient Inference with Qwen3.5-4BEfficient LLM inference and serving → Speculative decoding
Progress- and Reliability-Oriented Group Policy Optimization for Agentic Reinforcement LearningLLM agents and tool use → Reinforcement learning for LLM agents
Hierarchical Multi-to-Single-Modal Knowledge Distillation for Disruption Prediction in EASTKnowledge distillation → General knowledge distillation methods
Biological Motifs for Agentic ControlLLM agents and tool use → General LLM agent frameworks and benchmarks
Spinning Straw into Gold: Relabeling LLM Agent Trajectories in Hindsight for Successful DemonstrationsLLM agents and tool use → Reinforcement learning for LLM agents
Teaching Code LLMs to Reason with Intermediate Formal SpecificationsLLM reasoning → Reasoning for agents, planning and code
Detecting Hallucinations in Retrieval-Augmented Generation through Grounding-Aware Sensitivity by Perturbation (GASP)Retrieval-augmented generation → Hallucination detection in RAG
Agentic IoT: Architectures, Applications, and Challenges Toward the Internet of AgentsLLM agents and tool use → Agentic AI for wireless networks
An Evaluation of Role-Based Multi-Agent Code Generation on Repository-Scale ProblemsLLM agents and tool use → Software engineering coding agents
Sangam: Efficiently Serving Diffusion LLMs with the AR StackDiffusion models → Discrete and text diffusion
CoCoScale: Leveraging Layer-wise Scaling to Unlock the Potential of Online LLM ServingEfficient LLM inference and serving → Serving systems and request scheduling
CritiqueDriveVLM: From Verifier-Guided Reinforcement Learning to Latent Thought Distillation for Autonomous DrivingLLM reasoning → Chain-of-thought reasoning for driving +3
Teaching Tiny VLA Models Where to Look and How to MoveVision-language-action models → Generalist VLA models and benchmarks
BrownoutMoE: Structure-Aware Expert Grouping for Efficient and Accurate LLM Web-based ServicesMixture-of-experts → Efficient MoE training and inference systems
SeeMe: Mitigating Hallucinations in Large Vision-Language Models through Effective Visual Token EngineeringMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
ACE: Agentic Control for Embodied Manipulation via Zero-shot Workflow ReasoningLLM agents and tool use → Multimodal and embodied agents
Mask-based Predictive Representations for Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
Beyond Scene Priors: Fine-Grained Traffic Scene Reasoning with Benchmarking and Query-Guided Small-Object FocusLLM reasoning → Chain-of-thought reasoning for driving
!Imperio, smolVLA: The Implications of Data Poisoning on Open Source RoboticsLLM safety and security → Prompt injection and agent security
DELTA-TTS: Adapting Autoregressive Model into Diffusion Language Model for Text-to-SpeechDiffusion models → Audio, speech and music
SOV-CAD: Stepwise Orthographic Views Guided CAD Modeling Sequence ReconstructionOffline RL and imitation learning from data → Robot manipulation imitation learning
Asymptotic Preservation and Uniform Accuracy of Diffusion and Flow-Matching SamplersDiffusion models → Flow matching, time series and science
DynaVieW: Schema-Guided World Modeling for Understanding Hierarchical Visual DynamicsWorld models → Video-generation world models and simulators +1
Forethought: Verifiable Reasoning from Neurosymbolic Primitive ProgrammingLLM reasoning → Reasoning benchmarks and evaluation +1
Target-Aware Interaction-Guided Reinforcement Learning for Black-Box Node Injection Attacks on Graph Neural NetworksDeep reinforcement learning methods → Adversarial robustness and RL in finance
Seeing Once is Enough? Online Geometry-Aware Token Pruning for 3D Question AnsweringMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
DisclosureBeta: A Measurement-Channel Theory for Regime-Conditioned Betas from LLM-Read Risk Disclosures
Contamination Inflates Scores but Rarely Reorders Large Language Model Leaderboards
Discrete-Time MDP Modeling for Multi-Item Capacitated Lot Sizing with Stochastic Demand Timing
Interactive Analysis of Global Explanations using Aggregated Class Activation Maps for Network Data
New Terms, New Toxicity: Consensus-based Chinese Neologism Toxicity Detection via Search-Augmented LLMs
PsiLogic: Chaos-Aware Active Cancellation for Adam with a Fair Cross-Domain Benchmark
Benchmarking Sensor Robustness in Plasma Diagnostic Models: A Systematic Evaluation on TokaMark
Burst Spiking Neural Networks
RSRA: Training-Free Probing of Representation Sensitivity for Efficient LoRA Rank Allocation
LegalFarePlan: A Label-Setting Framework for Fare-Transparent Urban Rail Route Planning under Non-Additive Fare Rules
Cross-Subject Modeling for Widefield Calcium Imaging via Atlas-Aligned Spatiotemporal Tokenization
CoFINN: Conservation Flux Informed Neural Networks for Physics Problems Governed by Conservation Laws
Pixel-Precise Explainable Stress Indexing: A Semantic Segmentation Framework for Disease Severity Quantification in Field Crops
Empirical Minimal-Realisation Compression of Deep Neural Networks via Controllability-Observability Tests
Exogenous Dropout: A Simple, Strong Baseline for Corruption-Robust Time Series Forecasting with Covariates
The Granularity Paradox: How Temporal Disaggregation Inflates In-Sample Fit and Compounds Out-of-Sample Error
Geometry-Aware Infrastructure-Anchored Denoiser for UWB Sensing and Work-Zone Reconstruction
Lights, Camera, Carbon: Architectural Scaling Laws for Video Generation Energy Consumption
Explainable Novel Category Discovery in Semantic Concept Space
SceneFrom3D: Geometry-Conditioned Outdoor 3D Scene Generation via View Scheduling with Object-Level Control
Obey, Diverge, Collapse: Blind Obedience to Incorrect Instructions Drives Code LLMs to Irrecoverable Code Semantic Collapse
ManifoldFlow: Spectrum-Learnable Stiefel Optimization via SPD Relaxation
Mechanism-level routing failure in LLMs over Lean-verified algebraic structures
Lyapunov-Guided Training for Hardware-Safe Neural Networks Under Fixed-Point Arithmetic
Causal ASCEND: Scalable Two-tier Causal Discovery on High Dimensional Multi-omics Data

All 211