Papers

New on arXiv

Mon 17 Feb 2025

484 papers

What Are They Filtering Out? An Experimental Benchmark of Filtering Strategies for Harm Reduction in Pretraining DatasetsLLM safety and security → Safety alignment, refusal and guardrails
Invisible Walls in Cities: Designing LLM Agent to Predict Urban Segregation Experience with Social Media ContentLLM agents and tool use → Social simulation and multi-agent debate
Rotate, Clip, and Partition: Towards W2A4KV4 Quantization by Integrating Rotation and Learnable Non-uniform QuantizerEfficient LLM inference and serving → KV cache quantization
A Comprehensive Survey on Concept Erasure in Text-to-Image Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects
CoDiff: Conditional Diffusion Model for Collaborative 3D Object DetectionDiffusion models → Image generation, text-to-image and editing
Thinking Preference OptimizationLLM reasoning → Test-time compute and efficient thinking +1
Unveiling Privacy Risks in LLM Agent MemoryLLM agents and tool use → Agent threats, guardrails and governance
Unveiling the Magic of Code Reasoning through Hypothesis Decomposition and AmendmentLLM reasoning → Reasoning for agents, planning and code
HedgeAgents: A Balanced-aware Multi-agent Financial Trading SystemLLM agents and tool use → Financial trading and forecasting agents
Locally-Deployed Chain-of-Thought (CoT) Reasoning Model in Chemical Engineering: Starting from 30 Experimental DataLLM reasoning → Clinical and biomedical applications
UltraGen: Extremely Fine-grained Controllable Generation via Attribute Reconstruction and Global Preference OptimizationLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
IMLE Policy: Fast and Sample Efficient Visuomotor Policy Learning via Implicit Maximum Likelihood EstimationDiffusion models → Diffusion and flow policies for robotics +1
LanP: Rethinking the Impact of Language Priors in Large Vision-Language ModelsLLM reasoning → Hallucination detection and mitigation +1
Hovering Flight of Soft-Actuated Insect-Scale Micro Aerial Vehicles using Deep Reinforcement LearningDeep reinforcement learning methods → RL for robot navigation and UAV control
Towards Mechanistic Interpretability of Graph Transformers via Attention GraphsMechanistic interpretability → Circuits and mechanistic analysis
REAL-MM-RAG: A Real-World Multi-Modal Retrieval BenchmarkRetrieval-augmented generation → RAG pipeline design and evaluation
X-IL: Exploring the Design Space of Imitation Learning PoliciesOffline RL and imitation learning from data → Imitation learning methods
Towards Fusing Point Cloud and Visual Representations for Imitation LearningOffline RL and imitation learning from data → Robot manipulation imitation learning
Evaluating Step-by-step Reasoning Traces: A SurveyLLM reasoning → Reasoning benchmarks and evaluation
Connecting Large Language Model Agent to High Performance Computing ResourceLLM agents and tool use → Web, search, and research agents
Integrating Expert Knowledge into Logical Programs via LLMsLLM reasoning → Reasoning benchmarks and evaluation
Learning to Reason at the Frontier of LearnabilityLLM reasoning → Reinforcement learning for reasoning
InfoQuest: Evaluating Multi-Turn Dialogue Agents for Open-Ended Conversations with Hidden ContextLLM agents and tool use → General LLM agent frameworks and benchmarks
Fate: Fast Edge Inference of Mixture-of-Experts Models via Cross-Layer GateMixture-of-experts → Efficient MoE training and inference systems
Tactic: Adaptive Sparse Attention with Clustering and Distribution Fitting for Long-Context LLMsEfficient LLM inference and serving → KV cache optimization
Revisiting the Test-Time Scaling of o1-like Models: Do they Truly Possess Test-Time Scaling Capabilities?LLM reasoning → Test-time compute and efficient thinking
Zero Token-Driven Deep Thinking in LLMs: Unlocking the Full Potential of Existing Parameters via Cyclic RefinementLLM reasoning → Test-time compute and efficient thinking +1
Enhancing Frame Detection with Retrieval Augmented GenerationRetrieval-augmented generation → Multimodal RAG
Diffusion Models without Classifier-free GuidanceDiffusion models → Diffusion foundations and sampling
HARBOR: Exploring Persona Dynamics in Multi-Agent CompetitionLLM agents and tool use → Financial trading and forecasting agents
HermesFlow: Seamlessly Closing the Gap in Multimodal Understanding and GenerationMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Diffusion-Sharpening: Fine-tuning Diffusion Models with Denoising Trajectory SharpeningDiffusion models → Diffusion and flow policies for robotics
Fast or Better? Balancing Accuracy and Cost in Retrieval-Augmented Generation with Flexible User ControlRetrieval-augmented generation → RAG pipeline design and evaluation
Small Models Struggle to Learn from Strong ReasonersLLM reasoning → Prompting and chain-of-thought reasoning
SoftCoT: Soft Chain-of-Thought for Efficient Reasoning with LLMsLLM reasoning → Prompting and chain-of-thought reasoning
Transformer Dynamics: A neuroscientific approach to interpretability of large language modelsMechanistic interpretability → Circuits and mechanistic analysis
Scaling Autonomous Agents via Automatic Reward Modeling And PlanningLLM reasoning → Reasoning for agents, planning and code +2
RA-MTR: A Retrieval Augmented Multi-Task Reader based Approach for Inspirational Quote Extraction from Long DocumentsRetrieval-augmented generation → RAG robustness and adaptive retrieval
On the Query Complexity of Verifier-Assisted Language GenerationLLM reasoning → Test-time compute and efficient thinking
PRISM: Self-Pruning Intrinsic Selection Method for Training-Free Multimodal Data SelectionMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Scaling Test-Time Compute Without Verification or RL is SuboptimalLLM reasoning → Test-time compute and efficient thinking
A-MEM: Agentic Memory for LLM AgentsLLM agents and tool use → Agent memory systems
Descriminative-Generative Custom Tokens for Vision-Language ModelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
A Study on Leveraging Search and Self-Feedback for Agent ReasoningLLM reasoning → Reasoning for agents, planning and code
How Compositional Generalization and Creativity Improve as Diffusion Models are TrainedDiffusion models → Diffusion foundations and sampling
HumanGif: Single-View Human Diffusion with Generative PriorDiffusion models → 3D, point cloud and scene generation
TokenSkip: Controllable Chain-of-Thought Compression in LLMsLLM reasoning → Test-time compute and efficient thinking
PhysReason: A Comprehensive Benchmark towards Physics-Based ReasoningLLM reasoning → Spatial and embodied reasoning
A Survey on Bridging EEG Signals and Generative AI: From Image and Text to BeyondDiffusion models → Medical imaging and segmentation
KnowPath: Knowledge-enhanced Reasoning via LLM-generated Inference Paths over Knowledge GraphsLLM reasoning → Knowledge- and tool-augmented reasoning
SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning CapabilitiesLLM reasoning → Reasoning for agents, planning and code
Teaching LLMs According to Their Aptitude: Adaptive Reasoning for Mathematical Problem SolvingLLM reasoning → Mathematical reasoning +1
Atom of Thoughts for Markov LLM Test-Time ScalingLLM reasoning → Test-time compute and efficient thinking
Characterizing Photorealism and Artifacts in Diffusion Model-Generated ImagesDiffusion models → Image generation, text-to-image and editing
Image Inversion: A Survey from GANs to Diffusion and BeyondDiffusion models → Diffusion foundations and sampling
Learning Generalizable Prompt for CLIP with Class Similarity KnowledgeMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Theoretical Barriers in Bellman-Based Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
STRIVE: Structured Reasoning for Self-Improvement in Claim VerificationLLM reasoning → Reasoning benchmarks and evaluation
Massively Scaling Explicit Policy-conditioned Value FunctionsDeep reinforcement learning methods → Deep RL algorithms and representations
FitLight: Federated Imitation Learning for Plug-and-Play Autonomous Traffic Signal ControlOffline RL and imitation learning from data → Imitation and offline RL for driving +1
GRAPHGPT-O: Synergistic Multimodal Comprehension and Generation on GraphsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
EssayJudge: A Multi-Granular Benchmark for Assessing Automated Essay Scoring Capabilities of Multimodal Large Language ModelsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Neural Guided Diffusion BridgesDiffusion models → Diffusion foundations and sampling
Exploring Quantum Control Landscape and Solution Space Complexity through Dimensionality Reduction & Optimization AlgorithmsDeep reinforcement learning methods → Quantum reinforcement learning
MMRC: A Large-Scale Benchmark for Understanding Multimodal Large Language Model in Real-World ConversationMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Building A Proof-Oriented Programmer That Is 64% Better Than GPT-4o Under Data ScarcityLLM reasoning → Mathematical reasoning
LIMR: Less is More for RL ScalingLLM reasoning → Reinforcement learning for reasoning
Leveraging Dual Process Theory in Language Agent Framework for Real-time Simultaneous Human-AI CollaborationLLM agents and tool use → General LLM agent frameworks and benchmarks
Hypothesis-Driven Theory-of-Mind Reasoning for Large Language ModelsLLM reasoning → Reasoning benchmarks and evaluation
Does Knowledge About Perceptual Uncertainty Help an Agent in Automated Driving?End-to-end autonomous driving → Reinforcement learning for driving
Defining and Evaluating Visual Language Models' Basic Spatial Abilities: A Perspective from PsychometricsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
StructTransform: A Scalable Attack Surface for Safety-Aligned Large Language ModelsLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks
Can LLM Agents Maintain a Persona in Discourse?LLM agents and tool use → Social simulation and multi-agent debate
Code-Vision: Evaluating Multimodal LLMs Logic Understanding and Code Generation CapabilitiesMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Intersectional Fairness in Reinforcement Learning with Large State and Constraint SpacesDeep reinforcement learning methods → RL for games and combinatorial optimization
Less is More: Compact Clue Selection for Efficient Retrieval-Augmented Generation ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning +1
Table-Critic: A Multi-Agent Framework for Collaborative Criticism and Refinement in Table ReasoningLLM reasoning → Reasoning for agents, planning and code +1
BackdoorDM: A Comprehensive Benchmark for Backdoor Learning on Diffusion ModelDiffusion models → Robustness, privacy and adversarial aspects +1
video-SALMONN-o1: Reasoning-enhanced Audio-visual Large Language ModelLLM reasoning → Audio-language model reasoning
Cognitive-Aligned Document Selection for Retrieval-augmented GenerationRetrieval-augmented generation → RAG pipeline design and evaluation
Warmup-Distill: Bridge the Distribution Mismatch between Teacher and Student before Knowledge DistillationKnowledge distillation → Knowledge distillation for LLMs
Language Models Can See Better: Visual Contrastive Decoding For LLM Multimodal ReasoningLLM reasoning → Multimodal reasoning
ILIAS: Instance-Level Image retrieval At ScaleMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
FUNCTO: Function-Centric One-Shot Imitation Learning for Tool ManipulationOffline RL and imitation learning from data → Robot manipulation imitation learning
SQL-o1: A Self-Reward Heuristic Dynamic Search Method for Text-to-SQLSynthetic data and self-improvement for LLMs → Synthetic reasoning data and bootstrapping
Mitigating Visual Knowledge Forgetting in MLLM Instruction-tuning via Modality-decoupled Gradient DescentLLM reasoning → Multimodal reasoning
MT-RAIG: Novel Benchmark and Evaluation Framework for Retrieval-Augmented Insight Generation over Multiple TablesRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
"See the World, Discover Knowledge": A Chinese Factuality Evaluation for Large Vision Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
LLM Agents Making Agent ToolsLLM agents and tool use → Software engineering coding agents
MVTokenFlow: High-quality 4D Content Generation using Multiview Token FlowDiffusion models → 3D, point cloud and scene generation
Improve LLM-as-a-Judge Ability as a General AbilityLLM reasoning → Reasoning benchmarks and evaluation +1
MathFimer: Enhancing Mathematical Reasoning by Expanding Reasoning Steps through Fill-in-the-Middle TaskLLM reasoning → Mathematical reasoning
RIDE: Enhancing Large Language Model Alignment through Restyled In-Context Learning Demonstration ExemplarsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
MaskGWM: A Generalizable Driving World Model with Video Mask ReconstructionWorld models → World models for autonomous driving +1
Uncovering the Impact of Chain-of-Thought Reasoning for Direct Preference Optimization: Lessons from Text-to-SQLLLM reasoning → Mathematical reasoning +1
MMXU: A Multi-Modal and Multi-X-ray Understanding Dataset for Disease ProgressionMultimodal LLMs / vision-language models → Medical vision-language models
Competing LLM Agents in a Non-Cooperative Game of Opinion PolarisationLLM agents and tool use → Social simulation and multi-agent debate
DELMAN: Dynamic Defense Against Large Language Model Jailbreaking with Model EditingLLM safety and security → Jailbreaking
Membership Inference Attacks for Face Images Against Fine-Tuned Latent Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects +1
Maximum Entropy Reinforcement Learning with Diffusion PolicyDiffusion models → Diffusion and flow policies for robotics +1

All 484