Papers

New on arXiv

Sun 18 May 2025

260 papers

KVmix: Gradient-Based Layer Importance-Aware Mixed-Precision Quantization for KV CacheEfficient LLM inference and serving → KV cache quantization
How malicious AI swarms can threaten democracy: The fusion of agentic AI and LLMs marks a new frontier in information warfareLLM agents and tool use → Agent threats, guardrails and governance
Discovering and Steering Interpretable Concepts in Large Generative Music ModelsMechanistic interpretability → Activation steering methods
Improving LLM Outputs Against Jailbreak Attacks with Expert Model IntegrationLLM safety and security → Jailbreaking
Synthetic History: Evaluating Visual Representations of the Past in Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
Synthetic Data RL: Task Definition Is All You NeedSynthetic data and self-improvement for LLMs → Synthetic reasoning data and bootstrapping +1
SPIRIT: Patching Speech Language Models against Jailbreak AttacksLLM safety and security → Audio jailbreaks and speech backdoors
RAGXplain: From Explainable Evaluation to Actionable Guidance of RAG PipelinesRetrieval-augmented generation → RAG pipeline design and evaluation
Distributional Soft Actor-Critic with Harmonic Gradient for Safe and Efficient Autonomous Driving in Multi-lane ScenariosEnd-to-end autonomous driving → Reinforcement learning for driving +1
BARREL: Boundary-Aware Reasoning for Factual and Reliable LRMsLLM reasoning → Test-time compute and efficient thinking
Logic Jailbreak: Efficiently Unlocking LLM Safety Restrictions Through Formal Logical ExpressionLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
RealMath: A Continuous Benchmark for Evaluating Language Models on Research-Level MathematicsLLM reasoning → Mathematical reasoning
Uncovering Competing Poisoning Attacks in Retrieval-Augmented GenerationRetrieval-augmented generation → Privacy and security of RAG
Batched Self-Consistency Improves LLM Relevance Assessment and RankingLLM reasoning → Knowledge- and tool-augmented reasoning
A Survey of Attacks on Large Language ModelsLLM agents and tool use → Agent threats, guardrails and governance
Alternators With Noise ModelsDiffusion models → Flow matching, time series and science
Disambiguation in Conversational Question Answering in the Era of LLMs and Agents: A SurveyLLM agents and tool use → General LLM agent frameworks and benchmarks
ESC-Judge: A Framework for Comparing Emotional Support Conversational AgentsLLM agents and tool use → Medical and scientific domain agents
Reasoning by Superposition: A Theoretical Perspective on Chain of Continuous ThoughtLLM reasoning → Prompting and chain-of-thought reasoning
Revitalizing Black-Box Interpretability: Actionable Interpretability for LLMs via Proxy ModelsMechanistic interpretability → Knowledge and representations inside LLMs
InnateCoder: Learning Programmatic Options with Foundation ModelsDeep reinforcement learning methods → RL training of language and agent models
CPGD: Toward Stable Rule-based Reinforcement Learning for Language ModelsLLM reasoning → Reinforcement learning for reasoning
ALAS: A Stateful Multi-LLM Agent Framework for Disruption-Aware PlanningLLM agents and tool use → General LLM agent frameworks and benchmarks
MARGE: Improving Math Reasoning for LLMs with Guided ExplorationLLM reasoning → Mathematical reasoning
GUI-Shift: Enhancing VLM-Based GUI Agents through Self-supervised Reinforcement LearningMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
Video-GPT via Next Clip DiffusionWorld models → Video-generation world models and simulators
Guiding Diffusion with Deep Geometric Moments: Balancing Fidelity and VariationDiffusion models → 3D, point cloud and scene generation
Enhancing Large Language Models with Reward-guided Tree Search for Knowledge Graph Question and AnsweringLLM reasoning → Knowledge- and tool-augmented reasoning +1
Beyond Frameworks: Unpacking Collaboration Strategies in Multi-Agent SystemsLLM agents and tool use → General LLM agent frameworks and benchmarks
Resolving Latency and Inventory Risk in Market Making with Reinforcement LearningDeep reinforcement learning methods → Adversarial robustness and RL in finance
Model-Free Robust Average-Reward Reinforcement Learning with Sample Complexity AnalysisDeep reinforcement learning methods → RL theory and sample complexity
UFO-RL: Uncertainty-Focused Optimization for Efficient Reinforcement Learning Data SelectionLLM reasoning → Reinforcement learning for reasoning +1
SSR: Enhancing Depth Perception in Vision-Language Models via Rationale-Guided Spatial ReasoningLLM reasoning → Spatial and embodied reasoning +1
BadNAVer: Exploring Jailbreak Attacks On Vision-and-Language NavigationLLM reasoning → Reasoning for agents, planning and code +1
IP Leakage Attacks Targeting LLM-Based Multi-Agent SystemsLLM agents and tool use → Agent threats, guardrails and governance
SGDPO: Self-Guided Direct Preference Optimization for Language Model AlignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
VideoRFT: Incentivizing Video Reasoning Capability in MLLMs via Reinforced Fine-TuningLLM reasoning → Multimodal reasoning
Observe-R1: Unlocking Reasoning Abilities of MLLMs with Dynamic Progressive Reinforcement LearningLLM reasoning → Multimodal reasoning
Depth Transfer: Learning to See Like a Simulator for Real-World Drone NavigationDeep reinforcement learning methods → RL for robot navigation and UAV control
DragLoRA: Online Optimization of LoRA Adapters for Drag-based Image Editing in Diffusion ModelDiffusion models → Image generation, text-to-image and editing
Table-R1: Region-based Reinforcement Learning for Table UnderstandingLLM reasoning → Knowledge- and tool-augmented reasoning +1
Training Latent Diffusion Models with Interacting Particle AlgorithmsDiffusion models → Diffusion foundations and sampling
MTIL: Encoding Full History with Mamba for Temporal Imitation LearningWorld models → Embodied and robot world models +2
Automated Profile Inference with Language Model AgentsLLM agents and tool use → Agent threats, guardrails and governance
Traversal Verification for Speculative Tree DecodingEfficient LLM inference and serving → Speculative decoding
LLM-CoT Enhanced Graph Neural Recommendation with Harmonized Group Policy OptimizationLLM reasoning → LLM reasoning for recommendation
Few-Shot Concept Unlearning with Low Rank AdaptationDiffusion models → Robustness, privacy and adversarial aspects +1
Generative Diffusion Model Driven Massive Random Access in Massive MIMO SystemsDiffusion models → Diffusion for compression and wireless
Graph-Reward-SQL: Execution-Free Reinforcement Learning for Text-to-SQL via Graph Matching and Stepwise RewardLLM alignment / RLHF / preference optimisation → Process reward models for reasoning +1
MedAgentBoard: Benchmarking Multi-Agent Collaboration with Conventional Methods for Diverse Medical TasksLLM agents and tool use → Medical and scientific domain agents
Enhancing Visual Grounding for GUI Agents via Self-Evolutionary Reinforcement LearningLLM agents and tool use → GUI and mobile agents +1
Fully Geometric Multi-Hop Reasoning on Knowledge Graphs with Transitive RelationsLLM reasoning → Knowledge- and tool-augmented reasoning
CAPTURE: Context-Aware Prompt Injection Testing and Robustness EnhancementLLM safety and security → Prompt injection and agent security
DisCO: Reinforcing Large Reasoning Models with Discriminative Constrained OptimizationLLM reasoning → Reinforcement learning for reasoning
Towards Visuospatial Cognition via Hierarchical Fusion of Visual ExpertsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
STAR: Stage-Wise Attention-Guided Token Reduction for Efficient Large Vision-Language Models InferenceMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
AbFlowNet: Optimizing Antibody-Antigen Binding Energy via Diffusion-GFlowNet FusionDiffusion models → Flow matching, time series and science
A universal policy wrapper with guaranteesDeep reinforcement learning methods → Safe and constrained RL
Multi-CALF: A Policy Combination Approach with Statistical GuaranteesDeep reinforcement learning methods → RL theory and sample complexity
Reasoning-CV: Fine-tuning Powerful Reasoning LLMs for Knowledge-Assisted Claim VerificationLLM reasoning → Reasoning benchmarks and evaluation
SEED-GRPO: Semantic Entropy Enhanced GRPO for Uncertainty-Aware Policy OptimizationLLM reasoning → Reinforcement learning for reasoning
Mitigating Hallucinations via Inter-Layer Consistency Aggregation in Large Vision-Language ModelsMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Is Artificial Intelligence Generated Image Detection a Solved Problem?Diffusion models → Image generation, text-to-image and editing
VoiceCloak: A Multi-Dimensional Defense Framework against Unauthorized Diffusion-based Voice CloningDiffusion models → Audio, speech and music
LLMSR@XLLM25: An Empirical Study of LLM for Structural ReasoningLLM reasoning → Reasoning benchmarks and evaluation
Robust Planning for Autonomous Driving via Mixed Adversarial Diffusion PredictionsDiffusion models → Diffusion and flow policies for robotics +1
Model alignment using inter-modal bridgesLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
BeliefNest: A Joint Action Simulator for Embodied Agents with Theory of MindLLM agents and tool use → Reinforcement learning for LLM agents
ExpertSteer: Intervening in LLMs through Expert KnowledgeMechanistic interpretability → Activation steering methods
Visuospatial Cognitive AssistantMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
LogicOCR: Do Your Large Multimodal Models Excel at Logical Reasoning on Text-Rich Images?LLM reasoning → Multimodal reasoning +1
MobileIPL: Enhancing Mobile Agents Thinking Process via Iterative Preference LearningLLM reasoning → Multimodal reasoning +2
The Tower of Babel Revisited: Multilingual Jailbreak Prompts on Closed-Source Large Language ModelsLLM safety and security → Jailbreaking
Shorten After You're Right: Lazy Length Penalties for Reasoning RLLLM reasoning → Test-time compute and efficient thinking
Addressing Missing Data Issue for Diffusion-based RecommendationDiffusion models → Diffusion recommenders
Emergent Active Perception and Dexterity of Simulated Humanoids from Visual Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
Curriculum Abductive LearningLLM reasoning → Reasoning benchmarks and evaluation
Context-Aware Autoregressive Models for Multi-Conditional Image GenerationDiffusion models → Image generation, text-to-image and editing
K-MSHC: Unmasking Minimally Sufficient Head Circuits in Large Language Models with Experiments on Syntactic Classification TasksMechanistic interpretability → Circuits and mechanistic analysis
LLaVA-4D: Embedding SpatioTemporal Prompt into LMMs for 4D Scene UnderstandingMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
LAMeTA: Intent-Aware Agentic Network Optimization via a Large AI Model-Empowered Two-Stage ApproachLLM agents and tool use → Reinforcement learning for LLM agents
Towards Efficient and Exact Forgetting Services in Pre-Trained-Model-based Continual LearningLLM safety and security → Machine unlearning
From Shots to Stories: LLM-Assisted Video Editing with Unified Language RepresentationsLLM reasoning → Multimodal reasoning +1
NOFT: Test-Time Noise Finetune via Information Bottleneck for Highly Correlated Asset CreationDiffusion models → Image generation, text-to-image and editing
Sentience Quest: Towards Embodied, Emotionally Adaptive, Self-Evolving, Ethically Aligned Artificial General IntelligenceSynthetic data and self-improvement for LLMs → Self-evolving LLM agents
Shallow Flow Matching for Coarse-to-Fine Text-to-Speech SynthesisDiffusion models → Audio, speech and music
Mining Intrinsic Rewards from LLM Hidden States for Efficient Best-of-N SamplingLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
RoboFAC: A Comprehensive Framework for Robotic Failure Analysis and CorrectionVision-language-action models → Failure recovery and robustness
Learning Impact-Rich Rotational Maneuvers via Centroidal Velocity Rewards and Sim-to-Real Techniques: A One-Leg Hopper Flip Case StudyDeep reinforcement learning methods → RL for robot locomotion and manipulation
HyperCap: Hyperspectral Land Cover Captioning Dataset for Vision Language ModelsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
GMSA: Enhancing Context Compression via Group Merging and Layer Semantic AlignmentEfficient LLM inference and serving → Prompt and context compression
Imagination-Limited Q-Learning for Offline Reinforcement LearningOffline RL and imitation learning from data → Offline RL algorithms
Of Mice and Machines: A Comparison of Learning Between Real World Mice and RL AgentsDeep reinforcement learning methods → RL for games and combinatorial optimization
Near-Optimal Sample Complexities of Divergence-based S-rectangular Distributionally Robust Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
Always Clear Depth: Robust Monocular Depth Estimation under Adverse WeatherDiffusion models → 3D, point cloud and scene generation
Spatial-LLaVA: Enhancing Large Language Models with Spatial Referring Expressions for Visual UnderstandingMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Mitigating Content Effects on Reasoning in Language Models through Fine-Grained Activation SteeringLLM reasoning → Test-time compute and efficient thinking +1
LLM-DSE: Searching Accelerator Parameters with LLM AgentsLLM agents and tool use → Software engineering coding agents
Self-Destructive Language ModelLLM safety and security → Safety alignment, refusal and guardrails
EVALOOOP: A Self-Consistency-Centered Framework for Assessing Large Language Model Robustness in ProgrammingLLM reasoning → Reasoning for agents, planning and code

All 260