Papers

New on arXiv

Thu 28 Aug 2025

346 papers

TinyServe: Query-Aware Cache Selection for Efficient LLM ServingEfficient LLM inference and serving → KV cache optimization
Emotionally-Aware Agents for Dispute ResolutionLLM agents and tool use → Social simulation and multi-agent debate
Adaptive Monitoring and Real-World Evaluation of Agentic AI SystemsLLM agents and tool use → Agent threats, guardrails and governance
AdaptCache: KV Cache Native Storage Hierarchy for Low-Delay and High-Quality Language Model ServingEfficient LLM inference and serving → KV cache optimization
Breaking the Cold-Start Barrier: Reinforcement Learning with Double and Dueling DQNsDeep reinforcement learning methods → RL training of language and agent models
RelP: Faithful and Efficient Circuit Discovery in Language Models via Relevance PatchingMechanistic interpretability → Circuits and mechanistic analysis
Weighted Support Points from Random Measures: An Interpretable Alternative for Generative ModelingDiffusion models → Diffusion foundations and sampling
Reinforcement Learning for Optimizing Large Qubit Array based Quantum Sensor CircuitsDeep reinforcement learning methods → Quantum reinforcement learning
A Mixture of Experts Gating Network for Enhanced Surrogate Modeling in External AerodynamicsMixture-of-experts → MoE theory and gating functions
Addressing accuracy and hallucination of LLMs in Alzheimer's disease research through knowledge graphsRetrieval-augmented generation → Medical and biomedical RAG
Decoding Memories: An Efficient Pipeline for Self-Consistency Hallucination DetectionLLM reasoning → Hallucination detection and mitigation
Designing Smarter Conversational Agents for Kids: Lessons from Cognitive Work and Means-Ends AnalysesLLM agents and tool use → General LLM agent frameworks and benchmarks
Improving Aviation Safety Analysis: Automated HFACS Classification Using Reinforcement Learning with Group Relative Policy OptimizationDeep reinforcement learning methods → Safe and constrained RL
FUTURE: Flexible Unlearning for Tree EnsembleLLM safety and security → Machine unlearning
WaveLLDM: Design and Development of a Lightweight Latent Diffusion Model for Speech Enhancement and RestorationDiffusion models → Audio, speech and music
A Survey of Scientific Large Language Models: From Data Foundations to Agent FrontiersLLM agents and tool use → Web, search, and research agents
The Percept-V Challenge: Can Multimodal LLMs Crack Simple Perception Problems?Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning
HiddenObject: Modality-Agnostic Fusion for Multimodal Hidden Object DetectionState-space sequence models (S4, Mamba) → Point clouds, detection and remote sensing
R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce LearningLLM reasoning → Reinforcement learning for reasoning +1
EO-1: An Open Unified Embodied Foundation Model for General Robot ControlLLM reasoning → Spatial and embodied reasoning
Automating the Deep Space Network Data Systems; A Case Study in Adaptive Anomaly Detection through Agentic AILLM agents and tool use → Web, search, and research agents
Learning to Generate Unit Test via Adversarial Reinforcement LearningDeep reinforcement learning methods → RL training of language and agent models
PVPO: Pre-Estimated Value-Based Policy Optimization for Agentic ReasoningLLM agents and tool use → Reinforcement learning for LLM agents
Beyond Prediction: Reinforcement Learning as the Defining Leap in Healthcare AIDeep reinforcement learning methods → RL for clinical treatment and surgery
Dress&Dance: Dress up and Dance as You Like It - Technical PreviewDiffusion models → 3D, point cloud and scene generation
OneReward: Unified Mask-Guided Image Generation via Multi-Task Human Preference LearningLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment +1
Language Models and Logic Programs for Trustworthy Tax ReasoningLLM reasoning → Reasoning for agents, planning and code
Veritas: Generalizable Deepfake Detection via Pattern-Aware ReasoningLLM reasoning → Multimodal reasoning
CogVLA: Cognition-Aligned Vision-Language-Action Model via Instruction-Driven Routing & SparsificationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
MMG-Vid: Maximizing Marginal Gains at Segment-level and Token-level for Efficient Video LLMsEfficient LLM inference and serving → Visual token pruning in multimodal models
Mitosis detection in domain shift scenarios: a Mamba-based approachState-space sequence models (S4, Mamba) → Mamba for medical image analysis
Reusing Computation in Text-to-Image Diffusion for Efficient Generation of Image SetsDiffusion models → Image generation, text-to-image and editing
An Agile Method for Implementing Retrieval Augmented Generation Tools in Industrial SMEsRetrieval-augmented generation → RAG pipeline design and evaluation
Phased One-Step Adversarial Equilibrium for Video Diffusion ModelsDiffusion models → Video generation
Inference-Time Alignment Control for Diffusion Models with Reinforcement Learning GuidanceDiffusion models → Diffusion and flow policies for robotics +1
ChainReaction: Causal Chain-Guided Reasoning for Modular and Explainable Causal-Why Video Question AnsweringLLM reasoning → Multimodal reasoning
Lethe: Purifying Backdoored Large Language Models with Knowledge DilutionLLM safety and security → Backdoor and data poisoning attacks
ProactiveEval: A Unified Evaluation Framework for Proactive Dialogue AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
DrivingGaussian++: Towards Realistic Reconstruction and Editable Simulation for Surrounding Dynamic Driving ScenesEnd-to-end autonomous driving → Generative simulation for driving
Guarding Against Malicious Biased Threats (GAMBiT) Experiments: Revealing Cognitive Bias in Human-Subjects Red-Team Cyber Range OperationsLLM safety and security → Prompt injection and agent security
How Can Input Reformulation Improve Tool Usage Accuracy in a Complex Dynamic Environment? A Study on τ-benchLLM agents and tool use → Tool learning and function calling
Learning Robust Spatial Representations from Binaural Audio through Feature DistillationKnowledge distillation → Speech and audio distillation
PromptSleuth: Detecting Prompt Injection via Semantic Intent InvarianceLLM safety and security → Prompt injection and agent security
MSRS: Evaluating Multi-Source Retrieval-Augmented GenerationRetrieval-augmented generation → RAG pipeline design and evaluation
AVIATOR: Towards AI-Agentic Vulnerability Injection Workflow for High-Fidelity, Large-Scale Code Security DatasetLLM agents and tool use → Agent threats, guardrails and governance
Misleading Large Language Models used (or misused) in Scientific Peer-Reviewing via Hidden Prompt-Injection AttacksLLM safety and security → Prompt injection and agent security
Leveraging Discriminative Latent Representations for Conditioning GAN-Based Speech EnhancementDiffusion models → Audio, speech and music
PathMR: Multimodal Visual Reasoning for Interpretable Pathology DiagnosisLLM reasoning → Clinical and biomedical applications
JADES: A Universal Framework for Jailbreak Assessment via Decompositional ScoringLLM safety and security → Jailbreaking
Learning Primitive Embodied World Models: Towards Scalable Robotic LearningWorld models → Embodied and robot world models
Estimating 2D Keypoints of Surgical Tools Using Vision-Language Models with Low-Rank AdaptationMultimodal LLMs / vision-language models → Medical vision-language models
cMALC-D: Contextual Multi-Agent LLM-Guided Curriculum Learning with Diversity-Based Context BlendingLLM agents and tool use → Reinforcement learning for LLM agents
Multi-Agent Penetration Testing AI for the WebLLM agents and tool use → Agent threats, guardrails and governance
Single Agent Robust Deep Reinforcement Learning for Bus Fleet ControlDeep reinforcement learning methods → RL for autonomous driving and traffic
Evaluating Compositional Generalisation in VLMs and Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
Unleashing Uncertainty: Efficient Machine Unlearning for Generative AIDiffusion models → Robustness, privacy and adversarial aspects +1
Turning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety InjectionLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
SKGE-SWIN: End-To-End Autonomous Vehicle Waypoint Prediction and Navigation Using Skip Stage Swin TransformerEnd-to-end autonomous driving → End-to-end system architecture
Occlusion Robustness of CLIP for Military Vehicle ClassificationMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
SeqVLM: Proposal-Guided Multi-View Sequences Reasoning via VLM for Zero-Shot 3D Visual GroundingLLM reasoning → Spatial and embodied reasoning
Provable Benefits of In-Tool Learning for Large Language ModelsLLM agents and tool use → Tool learning and function calling
Pref-GRPO: Pairwise Preference Reward-based GRPO for Stable Text-to-Image Reinforcement LearningLLM reasoning → Reinforcement learning for reasoning +1
Rethinking Testing for LLM Applications: Characteristics, Challenges, and a Lightweight Interaction ProtocolLLM reasoning → Reasoning for agents, planning and code
Re4: Scientific Computing Agent with Rewriting, Resolution, Review and RevisionLLM reasoning → Reasoning for agents, planning and code +1
rStar2-Agent: Agentic Reasoning Technical ReportLLM reasoning → Reinforcement learning for reasoning +1
Addressing Tokenization Inconsistency in Steganography and Watermarking Based on Large Language ModelsLLM safety and security → Watermarking generated content and models
EEGDM: Learning EEG Representation with Latent Diffusion ModelDiffusion models → Medical imaging and segmentation
MobileCLIP2: Improving Multi-Modal Reinforced TrainingKnowledge distillation → Multimodal distillation
Task Allocation for Autonomous Machines using Computational Intelligence and Deep Reinforcement LearningDeep reinforcement learning methods → RL for robot navigation and UAV control
Physics-Guided Residual Reinforcement Learning for Humanoid Narrow-Path TraversalDeep reinforcement learning methods → RL for robot locomotion and manipulation
Improving Alignment in LVLMs with Debiased Self-JudgmentMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
VarDiU: A Variational Diffusive Upper Bound for One-Step Diffusion DistillationDiffusion models → Diffusion foundations and sampling
CyberSleuth: Autonomous Blue-Team LLM Agent for Web Attack ForensicsLLM agents and tool use → Agent threats, guardrails and governance
CraftGraffiti: Exploring Human Identity with Custom Graffiti Art via Facial-Preserving Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
GDS Agent for Graph Algorithmic ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning +1
Improving the Accuracy of Amortized Model Comparison with Self-ConsistencyLLM reasoning → Test-time compute and efficient thinking
FastFit: Accelerating Multi-Reference Virtual Try-On via Cacheable Diffusion ModelsDiffusion models → 3D, point cloud and scene generation
Flowing Straighter with Conditional Flow Matching for Accurate Speech EnhancementDiffusion models → Audio, speech and music
A Graph Talks, But Who's Listening? Rethinking Evaluations for Graph-Language ModelsLLM reasoning → Knowledge- and tool-augmented reasoning
SUMMA: A Multimodal Large Language Model for Advertisement SummarizationMultimodal LLMs / vision-language models → Video understanding with MLLMs
Dyslexify: A Mechanistic Defense Against Typographic Attacks in CLIPMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
MedGR2: Breaking the Data Barrier for Medical Reasoning via Generative Reward LearningLLM reasoning → Clinical and biomedical applications +2
Enhancing Health Fact-Checking with LLM-Generated Synthetic DataSynthetic data and self-improvement for LLMs → Medical synthetic data and agents
MoTAS: MoE-Guided Feature Selection from TTS-Augmented Speech for Enhanced Multimodal Alzheimer's Early ScreeningMixture-of-experts → Medical and biomedical MoE
Video-MTR: Reinforced Multi-Turn Reasoning for Long Video UnderstandingLLM reasoning → Multimodal reasoning
QTMRL: An Agent for Quantitative Trading Decision-Making Based on Multi-Indicator Guided Reinforcement LearningDeep reinforcement learning methods → Adversarial robustness and RL in finance
Dual-Model Weight Selection and Self-Knowledge Distillation for Medical Image ClassificationKnowledge distillation → Distillation for medical imaging
MCP-Bench: Benchmarking Tool-Using LLM Agents with Complex Real-World Tasks via MCP ServersLLM agents and tool use → Tool learning and function calling
Towards Mitigating Excessive Forgetting in LLM Unlearning via Entanglement-Guidance with Proxy ConstraintLLM safety and security → Machine unlearning
Attacks on Approximate Caches in Text-to-Image Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects
KG-CQR: Leveraging Structured Relation Representations in Knowledge Graphs for Contextual Query RetrievalRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
MindGuard: Intrinsic Decision Inspection for Securing LLM Agents Against Metadata PoisoningLLM agents and tool use → Agent threats, guardrails and governance +1
Fact or Facsimile? Evaluating the Factual Robustness of Modern RetrieversRetrieval-augmented generation → RAG robustness and adaptive retrieval
AWorld: Orchestrating the Training Recipe for Agentic AILLM agents and tool use → Reinforcement learning for LLM agents
Measuring Reasoning Utility in LLMs via Conditional Entropy ReductionLLM reasoning → Test-time compute and efficient thinking
Uncertainty Under the Curve: A Sequence-Level Entropy Area Metric for Reasoning LLMLLM reasoning → Test-time compute and efficient thinking
Audio-Guided Visual Editing with Complex Multi-Modal PromptsDiffusion models → Audio, speech and music
Enhancing Mamba Decoder with Bidirectional Interaction in Multi-Task Dense PredictionState-space sequence models (S4, Mamba) → Vision Mamba backbones
NPG-Muse: Scaling Long Chain-of-Thought Reasoning with NP-Hard Graph ProblemsLLM reasoning → Prompting and chain-of-thought reasoning
Adaptive Root Cause Localization for Microservice Systems with Multi-Agent Recursion-of-ThoughtLLM reasoning → Reasoning for agents, planning and code +1

All 346