Papers

New on arXiv

Thu 11 Dec 2025

344 papers

SoDA: An Efficient Interaction Paradigm for the Agentic WebLLM agents and tool use → Web, search, and research agents
GLOW: Graph-Language Co-Encoding for Agentic Workflow Performance PredictionLLM agents and tool use → Software engineering coding agents
Zero-Knowledge Audit for Internet of Agents: Privacy-Preserving Communication Verification with Model Context ProtocolLLM agents and tool use → Agent threats, guardrails and governance
PyFi: Toward Pyramid-like Financial Image Understanding for VLMs via Adversarial AgentsMultimodal LLMs / vision-language models → Chart understanding and anomaly detection +1
Towards Accessible Physical AI: LoRA-Based Fine-Tuning of VLA Models for Real-World Robot ControlVision-language-action models → Generalist VLA models and benchmarks
CXL-SpecKV: A Disaggregated FPGA Speculative KV-Cache for Datacenter LLM ServingEfficient LLM inference and serving → KV cache optimization
Motif-2-12.7B-Reasoning: A Practitioner's Guide to RL Training RecipesLLM reasoning → Reinforcement learning for reasoning
CORL: Reinforcement Learning of MILP Policies Solved via Branch and BoundDeep reinforcement learning methods → Deep RL algorithms and representations
Image Tiling for High-Resolution Reasoning: Balancing Local Detail with Global ContextLLM reasoning → Multimodal reasoning
MiniScope: A Least Privilege Framework for Authorizing Tool Calling AgentsLLM agents and tool use → Agent threats, guardrails and governance
Automated Penetration Testing with LLM Agents and Classical PlanningLLM agents and tool use → Agent threats, guardrails and governance
Fast-FoundationStereo: Real-Time Zero-Shot Stereo MatchingKnowledge distillation → Detection and dense prediction distillation
Limits and Gains of Test-Time Scaling in Vision-Language ReasoningLLM reasoning → Multimodal reasoning +1
VGent: Visual Grounding via Modular Design for Disentangling Reasoning and PredictionLLM reasoning → Multimodal reasoning +1
Data-Driven Model Reduction using WeldNet: Windowed Encoders for Learning DynamicsMechanistic interpretability → Sparse autoencoder training and evaluation
VDAWorld: World Modelling via VLM-Directed Abstraction and SimulationWorld models → Video-generation world models and simulators
Synthetic Vasculature and Pathology Enhance Vision-Language Model ReasoningLLM reasoning → Clinical and biomedical applications +1
Weakly Supervised Tuberculosis Localization in Chest X-rays through Knowledge DistillationKnowledge distillation → Distillation for medical imaging
WorldLens: Full-Spectrum Evaluations of Driving World Models in Real WorldWorld models → Video-generation world models and simulators +1
Group Diffusion: Enhancing Image Generation by Unlocking Cross-Sample CollaborationDiffusion models → Diffusion foundations and sampling
Towards Efficient and Effective Multi-Camera Encoding for End-to-End DrivingEnd-to-end autonomous driving → VLMs for autonomous driving
ImplicitRDP: An End-to-End Visual-Force Diffusion Policy with Structural Slow-Fast LearningDiffusion models → Diffusion and flow policies for robotics
AlcheMinT: Fine-grained Temporal Control for Multi-Reference Consistent Video GenerationDiffusion models → Video generation
VL-JEPA: Joint Embedding Predictive Architecture for Vision-languageMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
OmniView: An All-Seeing Diffusion Model for 3D and 4D View SynthesisDiffusion models → 3D, point cloud and scene generation
Curriculum-Based Reinforcement Learning for Autonomous UAV Navigation in Unknown Curved Tubular ConduitDeep reinforcement learning methods → RL for robot navigation and UAV control
Asynchronous Reasoning: Training-Free Interactive Thinking LLMsLLM reasoning → Audio-language model reasoning
FoundationMotion: Auto-Labeling and Reasoning about Spatial Movement in VideosLLM reasoning → Spatial and embodied reasoning
Digital Twin Supervised Reinforcement Learning Framework for Autonomous Underwater NavigationDeep reinforcement learning methods → RL for robot navigation and UAV control
Multi-Granular Node Pruning for Causal Circuit DiscoveryMechanistic interpretability → Circuits and mechanistic analysis
DuetSVG: Unified Multimodal SVG Generation with Internal Visual GuidanceMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
Guided Transfer Learning for Discrete Diffusion ModelsDiffusion models → Discrete and text diffusion
From Macro to Micro: Benchmarking Microscopic Spatial Intelligence on Molecules via Vision-Language ModelsMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
Scaling Behavior of Discrete Diffusion Language ModelsDiffusion models → Discrete and text diffusion
UBCL: A Reinforcement Learning Framework for Controllable and Diverse Player BehaviorsDeep reinforcement learning methods → RL for games and combinatorial optimization
V-OCBF: Learning Safety Filters from Offline Data via Value-Guided Offline Control Barrier FunctionsOffline RL and imitation learning from data → Safe offline RL
Interpretable and Steerable Concept Bottleneck Sparse AutoencodersMechanistic interpretability → Sparse autoencoder training and evaluation
Replace, Don't Expand: Mitigating Context Dilution in Multi-Hop RAG via Fixed-Budget Evidence AssemblyRetrieval-augmented generation → Generation and context integration
OPV: Outcome-based Process Verifier for Efficient Long Chain-of-Thought VerificationLLM reasoning → Reinforcement learning for reasoning
Intern-S1-MO: Long-horizon Reasoning Agent for Olympiad?Level Mathematical Problem SolvingLLM reasoning → Mathematical reasoning
SpaceDrive: Infusing Spatial Awareness into VLM-based Autonomous DrivingLLM reasoning → Chain-of-thought reasoning for driving +2
How to Brake? Ethical Emergency Braking with Deep Reinforcement LearningDeep reinforcement learning methods → RL for autonomous driving and traffic
Remember Me, Refine Me: A Dynamic Procedural Memory Framework for Experience-Driven Agent EvolutionLLM agents and tool use → Agent memory systems
Enhancing Radiology Report Generation and Visual Grounding using Reinforcement LearningMultimodal LLMs / vision-language models → Medical vision-language models +1
Safe for Whom? Rethinking How We Evaluate the Safety of LLMs for Real UsersLLM reasoning → Reasoning for agents, planning and code
Evaluating Gemini Robotics Policies in a Veo World SimulatorWorld models → Embodied and robot world models
AEBNAS: Strengthening Exit Branches in Early-Exit Networks through Hardware-Aware Neural Architecture SearchEfficient LLM inference and serving → Early exit and edge inference
On the Dynamics of Multi-Agent LLM Communities Driven by Value DiversityLLM agents and tool use → Social simulation and multi-agent debate
Closing the Navigation Compliance Gap in End-to-end Autonomous DrivingEnd-to-end autonomous driving → End-to-end system architecture
CAPTAIN: Semantic Feature Injection for Memorization Mitigation in Text-to-Image Diffusion ModelsDiffusion models → Diffusion foundations and sampling
AgriGPT-Omni: A Unified Speech-Vision-Text Framework for Multilingual Agricultural IntelligenceMultimodal LLMs / vision-language models → Agricultural vision-language models
Thinking While Driving: A Concurrent Framework for Real-Time, LLM-Based Adaptive RoutingLLM reasoning → Reasoning for agents, planning and code
LEO-RobotAgent: A General-purpose Robotic Agent for Language-driven Embodied OperatorLLM agents and tool use → Multimodal and embodied agents
Multi-Objective Reward and Preference Optimization: Theory and AlgorithmsLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Unleashing Degradation-Carrying Features in Symmetric U-Net: Simpler and Stronger Baselines for All-in-One Image RestorationDiffusion models → Inverse problems and image restoration +1
RoleRMBench & RoleRM: Towards Reward Modeling for Profile-Based Role Play in Dialogue SystemsLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Causal Reasoning Favors Encoders: On The Limits of Decoder-Only ModelsLLM reasoning → Causal reasoning in LLMs
Grounding Everything in Tokens for Multimodal Large Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Blink: Dynamic Visual Token Resolution for Enhanced Multimodal UnderstandingMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Unlocking the Address Book: Dissecting the Sparse Semantic Structure of LLM Key-Value Caches via Sparse AutoencodersEfficient LLM inference and serving → KV cache optimization +1
Achieving Olympiad-Level Geometry Large Language Model Agent via Complexity Boosting Reinforcement LearningLLM agents and tool use → Educational agents and math problem solving
Inverse problems with diffusion models: MAP estimation via mode-seeking lossDiffusion models → Inverse problems and image restoration
Adaptive Replay Buffer for Offline-to-Online Reinforcement LearningOffline RL and imitation learning from data → Offline RL algorithms
Zero-shot 3D Map Generation with LLM Agents: A Dual-Agent Architecture for Procedural Content GenerationLLM agents and tool use → Multimodal and embodied agents
UACER: An Uncertainty-Adaptive Critic Ensemble Framework for Robust Adversarial Reinforcement LearningDeep reinforcement learning methods → Adversarial robustness and RL in finance
When Reject Turns into Accept: Quantifying the Vulnerability of LLM-Based Scientific Reviewers to Indirect Prompt InjectionLLM safety and security → Prompt injection and agent security
Clustered Federated Learning with Hierarchical Knowledge DistillationKnowledge distillation → Federated, robust and private distillation
Decoding Student Minds: Leveraging Conversational Agents for Psychological and Learning AnalysisLLM agents and tool use → Educational agents and math problem solving
Targeted Data Protection for Diffusion Model by Matching Training TrajectoryDiffusion models → Robustness, privacy and adversarial aspects
T-pro 2.0: An Efficient Russian Hybrid-Reasoning Model and PlaygroundLLM reasoning → Reasoning benchmarks and evaluation
Cooperative Retrieval-Augmented Generation for Question Answering: Mutual Information Exchange and Ranking by Contrasting LayersRetrieval-augmented generation → RAG robustness and adaptive retrieval
How to Trick Your AI TA: A Systematic Study of Academic Jailbreaking in LLM Code EvaluationLLM safety and security → Jailbreaking
Boosting RL-Based Visual Reasoning with Selective Adversarial Entropy InterventionLLM reasoning → Reinforcement learning for reasoning +1
Confucius Code Agent: Scalable Agent Scaffolding for Real-World CodebasesLLM agents and tool use → Software engineering coding agents
RoboNeuron: A Middle-Layer Infrastructure for Agent-Driven Orchestration in Embodied AIVision-language-action models → Failure recovery and robustness +1
Towards Fine-Grained Recognition with Large Visual Language Models: Benchmark and Optimization StrategiesMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Investigating training objective for flow matching-based speech enhancementDiffusion models → Audio, speech and music
AgentProg: Empowering Long-Horizon GUI Agents with Program-Guided Context ManagementLLM agents and tool use → GUI and mobile agents
GPG: Generalized Policy Gradient Theorem for Transformer-based PoliciesDeep reinforcement learning methods → RL theory and sample complexity
Visual Funnel: Resolving Contextual Blindness in Multimodal Large Language ModelsLLM reasoning → Multimodal reasoning +1
Tool-Augmented Spatiotemporal Reasoning for Streamlining Video Question Answering TaskLLM reasoning → Multimodal reasoning
Hybrid Transformer-Mamba for Weakly Supervised Volumetric Medical SegmentationState-space sequence models (S4, Mamba) → Mamba for medical image analysis
Geometric Dynamics of Agentic Loops in Large Language ModelsLLM reasoning → Test-time compute and efficient thinking +1
REMISVFU: Vertical Federated Unlearning via Representation Misdirection for Intermediate Output FeatureLLM safety and security → Federated unlearning and poisoning
CoSPlan: Corrective Sequential Planning via Scene Graph Incremental UpdatesMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
EchoingPixels: Aliasing-Resistant Joint Token Reduction for Audio-Visual LLMsEfficient LLM inference and serving → Audio and omni-modal token compression
EpiPlanAgent: Agentic Automated Epidemic Response PlanningLLM agents and tool use → Social simulation and multi-agent debate
Efficient-VLN: A Simple yet Strong Baseline for Efficient Vision-Language NavigationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Investigating The Functional Roles of Attention Heads in Vision Language Models: Evidence for Reasoning ModulesLLM reasoning → Multimodal reasoning +1
Reverse Thinking Enhances Missing Information Detection in Large Language ModelsLLM reasoning → Test-time compute and efficient thinking
MR-FlowDPO: Multi-Reward Direct Preference Optimization for Flow-Matching Text-to-Music GenerationDiffusion models → Audio, speech and music
RobustSora: De-Watermarked Benchmark for Robust AI-Generated Video DetectionLLM safety and security → Watermarking generated content and models
Multi-dimensional Preference Alignment by Conditioning Reward ItselfDiffusion models → Diffusion and flow policies for robotics +1
Task-Oriented Grasping Using Reinforcement Learning with a Contextual Reward MachineDeep reinforcement learning methods → RL for robot locomotion and manipulation
Latent Chain-of-Thought World Modeling for End-to-End DrivingWorld models → World models for autonomous driving +3
Does SWE-Bench-Verified Test Agent Ability or Model Memory?LLM agents and tool use → Software engineering coding agents
An exploration for higher efficiency in multi objective optimisation with reinforcement learningDeep reinforcement learning methods → RL for games and combinatorial optimization
AutoMedic: An Automated Evaluation Framework for Clinical Conversational Agents with Medical Dataset GroundingLLM agents and tool use → Medical and scientific domain agents
MINIF2F-DAFNY: LLM-Guided Mathematical Theorem Proving via Auto-Active VerificationLLM reasoning → Mathematical reasoning
Watermarks for Language Models via Probabilistic AutomataLLM safety and security → Watermarking generated content and models

All 344