Papers

New on arXiv

Tue 21 Oct 2025

376 papers

EdgeReasoning: Characterizing Reasoning LLM Deployment on Edge GPUsLLM reasoning → Reasoning for agents, planning and code
Activating Visual Context and Commonsense Reasoning through Masked Prediction in VLMsLLM reasoning → Multimodal reasoning
DiffGRM: Diffusion-based Generative Recommendation ModelDiffusion models → Diffusion recommenders
Can Reasoning Models Obfuscate Reasoning? Stress-Testing Chain-of-Thought MonitorabilityLLM reasoning → Prompting and chain-of-thought reasoning
Prompt Decorators: A Declarative and Composable Syntax for Reasoning, Formatting, and Control in LLMsLLM reasoning → Prompting and chain-of-thought reasoning
A Cross-Environment and Cross-Embodiment Path Planning Framework via a Conditional Diffusion ModelDiffusion models → Diffusion and flow policies for robotics
When Your AI Agent Succumbs to Peer-Pressure: Studying Opinion-Change Dynamics of LLMsLLM agents and tool use → Social simulation and multi-agent debate
What Makes a Good Curriculum? Disentangling the Effects of Data Ordering on LLM Mathematical ReasoningLLM reasoning → Mathematical reasoning
PoSh: Using Scene Graphs To Guide LLMs-as-a-Judge For Detailed Image DescriptionsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
The MUSE Benchmark: Probing Music Perception and Auditory Relational Reasoning in Audio LLMSLLM reasoning → Audio-language model reasoning +1
Rectifying Shortcut Behaviors in Preference-based Reward LearningLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Are they lovers or friends? Evaluating LLMs' Social Reasoning in English and Korean DialoguesLLM reasoning → Reasoning benchmarks and evaluation
MoAlign: Motion-Centric Representation Alignment for Video Diffusion ModelsDiffusion models → Video generation
XGen-Q: An Explainable Domain-Adaptive LLM Framework with Retrieval-Augmented Generation for Software SecurityRetrieval-augmented generation → Privacy and security of RAG
Robust Driving QA through Metadata-Grounded Context and Task-Specific PromptsLLM reasoning → Chain-of-thought reasoning for driving +2
Test-time Verification via Optimal Transport: Coverage, ROC, & Sub-optimalityLLM reasoning → Test-time compute and efficient thinking
Lost in the Maze: Overcoming Context Limitations in Long-Horizon Agentic SearchLLM agents and tool use → Web, search, and research agents
Noise-corrected GRPO: From Noisy Rewards to Unbiased GradientsLLM reasoning → Reinforcement learning for reasoning +1
Fairness Evaluation and Inference Level Mitigation in LLMsLLM safety and security → Safety alignment, refusal and guardrails
ADPO: Anchored Direct Preference OptimizationLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Learning from the Best, Differently: A Diversity-Driven Rethinking on Data SelectionLLM reasoning → Reasoning benchmarks and evaluation
Grasp Any Region: Towards Precise, Contextual Pixel Understanding for Multimodal LLMsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
EffiReasonTrans: RL-Optimized Reasoning for Code TranslationLLM reasoning → Reasoning benchmarks and evaluation
Every Step Evolves: Scaling Reinforcement Learning for Trillion-Scale Thinking ModelLLM reasoning → Reinforcement learning for reasoning
DP2O-SR: Direct Perceptual Preference Optimization for Real-World Image Super-ResolutionDiffusion models → Inverse problems and image restoration +1
Towards Faithful and Controllable Personalization via Critique-Post-Edit Reinforcement LearningLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Unifying and Enhancing Graph Transformers via a Hierarchical Mask FrameworkMixture-of-experts → MoE for time series and graphs
Search Self-play: Pushing the Frontier of Agent Capability without SupervisionLLM agents and tool use → Web, search, and research agents +1
Fine-Tuned Thoughts: Leveraging Chain-of-Thought Reasoning for Industrial Asset Health MonitoringLLM reasoning → Prompting and chain-of-thought reasoning
A Model Can Help Itself: Reward-Free Self-Training for LLM ReasoningLLM reasoning → Test-time compute and efficient thinking +1
KAT-Coder Technical ReportLLM agents and tool use → Software engineering coding agents
A Frequentist Statistical Introduction to Variational Inference, Autoencoders, and Diffusion ModelsDiffusion models → Diffusion foundations and sampling
HarmNet: A Framework for Adaptive Multi-Turn Jailbreak Attacks on Large Language ModelsLLM safety and security → Jailbreaking
IF-VidCap: Can Video Caption Models Follow Instructions?Multimodal LLMs / vision-language models → Video understanding with MLLMs
SSD: Spatial-Semantic Head Decoupling for Efficient Autoregressive Image GenerationEfficient LLM inference and serving → Fast autoregressive visual generation
Exploring a Unified Vision-Centric Contrastive Alternatives on Multi-Modal Web DocumentsMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
Fetch.ai: An Architecture for Modern Multi-Agent SystemsLLM agents and tool use → Agent threats, guardrails and governance
MoGA: Mixture-of-Groups Attention for End-to-End Long Video GenerationDiffusion models → Video generation
MLMA: Towards Multilingual ASR With Mamba-based ArchitecturesState-space sequence models (S4, Mamba) → Speech and audio SSMs
Exploring Membership Inference Vulnerabilities in Clinical Large Language ModelsLLM safety and security → Membership inference attacks
Reasoning Language Model Inference Serving Unveiled: An Empirical StudyLLM reasoning → Test-time compute and efficient thinking +1
Query Decomposition for RAG: Balancing Exploration-ExploitationRetrieval-augmented generation → RAG pipeline design and evaluation
Think with 3D: Geometric Imagination Grounded Spatial Reasoning from Limited ViewsLLM reasoning → Spatial and embodied reasoning +1
Visual Attention Reasoning via Hierarchical Search and Self-VerificationLLM reasoning → Multimodal reasoning +1
A Compositional Paradigm for Foundation Models: Towards Smarter Robotic AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
CUARewardBench: A Benchmark for Evaluating Reward Models on Computer-using AgentLLM agents and tool use → GUI and mobile agents +1
TokenCake: A KV-Cache-centric Serving Framework for LLM-based Multi-Agent ApplicationsEfficient LLM inference and serving → Efficient reasoning and agent inference
CLASP: Cost-Optimized LLM-based Agentic System for Phishing DetectionLLM agents and tool use → Agent threats, guardrails and governance
CovMatch: Cross-Covariance Guided Multimodal Dataset Distillation with Trainable Text EncoderMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
The Trust Paradox in LLM-Based Multi-Agent Systems: When Collaboration Becomes a Security VulnerabilityLLM agents and tool use → Agent threats, guardrails and governance
SOCIA-Nabla: Textual Gradient Meets Multi-Agent Orchestration for Automated Simulator GenerationLLM agents and tool use → Software engineering coding agents
Noise-Conditioned Mixture-of-Experts Framework for Robust Speaker VerificationMixture-of-experts → Multimodal MoE models
Counterfactual Reasoning for Steerable Pluralistic Value Alignment of Large Language ModelsLLM reasoning → Causal reasoning in LLMs +1
From Quarter to All: Accelerating Speculative LLM Decoding via Floating-Point Exponent Remapping and Parameter SharingEfficient LLM inference and serving → Speculative decoding
Socialized Learning and Emergent Behaviors in Multi-Agent Systems based on Multimodal Large Language ModelsLLM agents and tool use → Multimodal and embodied agents
Identity-Aware Large Language Models require Cultural ReasoningLLM reasoning → Reasoning benchmarks and evaluation
VAPU: System for Autonomous Legacy Code ModernizationLLM agents and tool use → Software engineering coding agents
Zero-Shot Vehicle Model Recognition via Text-Based Retrieval-Augmented GenerationRetrieval-augmented generation → Multimodal RAG
Crucible: Quantifying the Potential of Control Algorithms through LLM AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
AndroidControl-Curated: Revealing the True Potential of GUI Agents through Benchmark PurificationLLM agents and tool use → GUI and mobile agents
StarBench: A Turn-Based RPG Benchmark for Agentic Multimodal Decision-Making and Information SeekingLLM agents and tool use → Reinforcement learning for LLM agents
How Efficient Are Diffusion Language Models? A Critical Examination of Efficiency Evaluation PracticesDiffusion models → Discrete and text diffusion
LAFA: Agentic LLM-Driven Federated Analytics over Decentralized Data SourcesLLM agents and tool use → Web, search, and research agents
Probabilistic Modeling of Intentions in Socially Intelligent LLM AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
CircuitSeer: Mining High-Quality Data by Probing Mathematical Reasoning Circuits in LLMsLLM reasoning → Mathematical reasoning
VFM-VAE: Vision Foundation Models Can Be Good Tokenizers for Latent Diffusion ModelsDiffusion models → Diffusion foundations and sampling
ChronoPlay: A Framework for Modeling Dual Dynamics and Authenticity in Game RAG BenchmarksRetrieval-augmented generation → RAG pipeline design and evaluation
Engagement Undermines Safety: How Stereotypes and Toxicity Shape Humor in Language ModelsLLM safety and security → Safety alignment, refusal and guardrails
LAND: Lung and Nodule Diffusion for 3D Chest CT Synthesis with Anatomical GuidanceDiffusion models → Medical imaging and segmentation
PlanU: Large Language Model Reasoning through Planning under UncertaintyLLM reasoning → Reasoning for agents, planning and code
Grounding or Guessing? Visual Signals for Detecting Hallucinations in Sign Language TranslationMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Chain-of-Conceptual-Thought Elicits Daily Conversation in Large Language ModelsLLM reasoning → Prompting and chain-of-thought reasoning
AlphaOPT: Formulating Optimization Programs with Self-Improving LLM Experience LibrarySynthetic data and self-improvement for LLMs → Self-rewarding alignment and self-training
Med-VRAgent: A Framework for Medical Visual Reasoning-Enhanced AgentsLLM reasoning → Clinical and biomedical applications +2
Memory-Augmented State Machine Prompting: A Novel LLM Agent Framework for Real-Time Strategy GamesLLM agents and tool use → Reinforcement learning for LLM agents
Training Diverse Graph Experts for Ensembles: A Systematic Empirical StudyMixture-of-experts → MoE for time series and graphs
KoSimpleQA: A Korean Factuality Benchmark with an Analysis of Reasoning LLMsLLM reasoning → Reasoning benchmarks and evaluation
KrishokBondhu: A Retrieval-Augmented Voice-Based Agricultural Advisory Call Center for Bengali FarmersRetrieval-augmented generation → Time series and agricultural RAG
Ranking-based Preference Optimization for Diffusion Models from Implicit User FeedbackDiffusion models → Diffusion and flow policies for robotics +1
ECG-LLM -- training and evaluation of domain-specific large language models for electrocardiographyRetrieval-augmented generation → Medical and biomedical RAG
MoTVLA: A Vision-Language-Action Model with Unified Fast-Slow ReasoningVision-language-action models → Reasoning and chain-of-thought in VLA
Position: LLM Watermarking Should Align Stakeholders' Incentives for Practical AdoptionLLM safety and security → Watermarking generated content and models
Scalable, Explainable and Provably Robust Anomaly Detection with One-Step Flow MatchingDiffusion models → Flow matching, time series and science
Beyond Single Models: Mitigating Multimodal Hallucinations via Adaptive Token Ensemble DecodingMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
MoMaGen: Generating Demonstrations under Soft and Hard Constraints for Multi-Step Bimanual Mobile ManipulationOffline RL and imitation learning from data → Robot manipulation imitation learning
Higher Embedding Dimension Creates a Stronger World Model for a Simple Sorting TaskWorld models → World models for language agents
Genesis: Evolving Attack Strategies for LLM Web Agent Red-TeamingLLM agents and tool use → Agent threats, guardrails and governance +1
OmniNWM: Omniscient Driving Navigation World ModelsWorld models → World models for autonomous driving +1
The Impact of Image Resolution on Biomedical Multimodal Large Language ModelsMultimodal LLMs / vision-language models → Medical vision-language models
Proactive Reasoning-with-Retrieval Framework for Medical Multimodal Large Language ModelsLLM reasoning → Clinical and biomedical applications +1
GeoDiff: Geometry-Guided Diffusion for Metric Depth EstimationDiffusion models → 3D, point cloud and scene generation
Food4All: An Agentic Framework and Benchmark for Food Resource Navigation with Adaptive User UnderstandingLLM agents and tool use → Social simulation and multi-agent debate
Text or Pixels? It Takes Half: On the Token Efficiency of Visual Text Inputs in Multimodal LLMsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
StreamingTOM: Streaming Token Compression for Efficient Video UnderstandingMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
UWBench: A Comprehensive Vision-Language Benchmark for Underwater UnderstandingMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
DelvePO: Direction-Guided Self-Evolving Framework for Flexible Prompt OptimizationSynthetic data and self-improvement for LLMs → Self-rewarding alignment and self-training
Illusions of reflection: open-ended task reveals systematic failures in Large Language Models' reflective reasoningLLM reasoning → Test-time compute and efficient thinking
LIME: Link-based user-item Interaction Modeling with decoupled xor attention for Efficient test time scalingLLM reasoning → LLM reasoning for recommendation
RESCUE: Retrieval Augmented Secure Code GenerationRetrieval-augmented generation → Privacy and security of RAG
RadDiagSeg-M: A Vision Language Model for Joint Diagnosis and Multi-Target Segmentation in RadiologyMultimodal LLMs / vision-language models → Medical vision-language models

All 376