Papers

New on arXiv

Thu 20 Nov 2025

348 papers

The Linguistic Architecture of Reflective Thought: Evaluation of a Large Language Model as a Tool to Isolate the Formal Structure of MentalizationLLM reasoning → Reasoning benchmarks and evaluation
TeamPath: Building MultiModal Pathology Experts with Reasoning AI CopilotsLLM reasoning → Clinical and biomedical applications
MamTiff-CAD: Multi-Scale Latent Diffusion with Mamba+ for Complex Parametric SequenceDiffusion models → 3D, point cloud and scene generation +1
BlockCert: Certified Blockwise Extraction of Transformer MechanismsMechanistic interpretability → Circuits and mechanistic analysis
BOP-ASK: Object-Interaction Reasoning for Vision-Language ModelsLLM reasoning → Spatial and embodied reasoning +1
Towards Unified Vision Language Models for Forest Ecological Analysis in Earth ObservationLLM reasoning → Spatial and embodied reasoning +1
Jump-diffusion models of parametric volume-price distributionsDiffusion models → Flow matching, time series and science
Cognitive BASIC: An In-Model Interpreted Reasoning Language for LLMsLLM reasoning → Prompting and chain-of-thought reasoning
PEPPER: Perception-Guided Perturbation for Robust Backdoor Defense in Text-to-Image Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects +1
Stable diffusion models reveal a persisting human and AI gap in visual creativityDiffusion models → Image generation, text-to-image and editing
Membership Inference Attacks Beyond OverfittingLLM safety and security → Membership inference attacks
Revisiting Multimodal KV Cache Compression: A Frequency-Domain-Guided Outlier-KV-Aware ApproachMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
SafeR-CLIP: Mitigating NSFW Content in Vision-Language Models While Preserving Pre-Trained KnowledgeMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
AutoBackdoor: Automating Backdoor Attacks via LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance +1
Multi-Agent Code Verification via Information TheoryLLM agents and tool use → Software engineering coding agents
EvoLMM: Self-Evolving Large Multimodal Models with Continuous RewardsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning +1
Thinking-while-Generating: Interleaving Textual Reasoning throughout Visual GenerationLLM reasoning → Multimodal reasoning
Learning to Think Fast and Slow for Visual Language ModelsLLM reasoning → Multimodal reasoning +1
Video-as-Answer: Predict and Generate Next Video Event with Joint-GRPOLLM reasoning → Multimodal reasoning
V-ReasonBench: Toward Unified Reasoning Benchmark Suite for Video Generation ModelsLLM reasoning → Multimodal reasoning
Taming the Long-Tail: Efficient Reasoning RL Training with Adaptive DrafterLLM reasoning → Reinforcement learning for reasoning +1
Nemotron Elastic: Towards Efficient Many-in-One Reasoning LLMsLLM reasoning → Test-time compute and efficient thinking +2
Cognitive Foundations for Reasoning and Their Manifestation in LLMsLLM reasoning → Reasoning benchmarks and evaluation
Solving Spatial Supersensing Without Spatial SupersensingWorld models → Video-generation world models and simulators
Comparison of Text-Based and Image-Based Retrieval in Multimodal Retrieval Augmented Generation Large Language Model SystemsRetrieval-augmented generation → Multimodal RAG
Teacher-Guided One-Shot Pruning via Context-Aware Knowledge DistillationKnowledge distillation → General knowledge distillation methods
InternData-A1: Pioneering High-Fidelity Synthetic Data for Pre-training Generalist PolicyVision-language-action models → Generalist VLA models and benchmarks
TRIM: Scalable 3D Gaussian Diffusion Inference with Temporal and Spatial TrimmingDiffusion models → Diffusion foundations and sampling
SurvAgent: Hierarchical CoT-Enhanced Case Banking and Dichotomy-Based Multi-Agent System for Multimodal Survival PredictionLLM agents and tool use → Medical and scientific domain agents
Stabilizing Policy Gradient Methods via Reward ProfilingDeep reinforcement learning methods → RL theory and sample complexity
You Only Forward Once: An Efficient Compositional Judging ParadigmMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Time dependent loss reweighting for flow matching and diffusion models is theoretically justifiedDiffusion models → Flow matching, time series and science
TimeViper: A Hybrid Mamba-Transformer Vision-Language Model for Efficient Long Video UnderstandingMultimodal LLMs / vision-language models → Video understanding with MLLMs +1
D-GARA: A Dynamic Benchmarking Framework for GUI Agent Robustness in Real-World AnomaliesLLM agents and tool use → GUI and mobile agents
Erase to Retain: Low Rank Adaptation Guided Selective Unlearning in Medical Segmentation NetworksLLM safety and security → Machine unlearning
Boosting Predictive Performance on Tabular Data through Data Augmentation with Latent-Space Flow-Based DiffusionDiffusion models → Flow matching, time series and science
Contrastive vision-language learning with paraphrasing and negationMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Saving Foundation Flow-Matching Priors for Inverse ProblemsDiffusion models → Flow matching, time series and science
Flow and Depth Assisted Video Prediction with Latent TransformerWorld models → Video-generation world models and simulators
Limitations of Scalarisation in MORL: A Comparative Study in Discrete EnvironmentsDeep reinforcement learning methods → RL for games and combinatorial optimization
A Comparison Between Decision Transformers and Traditional Offline Reinforcement Learning AlgorithmsOffline RL and imitation learning from data → Decision transformers and sequence modeling
Anatomy of an Idiom: Tracing Non-Compositionality in Language ModelsMechanistic interpretability → Circuits and mechanistic analysis
LLaVA3: Representing 3D Scenes like a Cubist Painter to Boost 3D Scene Understanding of VLMsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Bridging the Semantic-Action Gap in Visual Token Pruning for Efficient VLA InferenceVision-language-action models → Efficient VLA inference and adaptation +1
FreqFlow: Long-term forecasting using lightweight flow matchingDiffusion models → Flow matching, time series and science
Second-Order MPC-Based Distributed Q-LearningDeep reinforcement learning methods → RL for autonomous driving and traffic
TOFA: Training-Free One-Shot Federated Adaptation for Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Trustworthy AI in the Agentic Lakehouse: from Concurrency to GovernanceLLM agents and tool use → Agent threats, guardrails and governance
An Agent-Based Framework for the Automatic Validation of Mathematical Optimization ModelsLLM agents and tool use → General LLM agent frameworks and benchmarks
The Shawshank Redemption of Embodied AI: Understanding and Benchmarking Indirect Environmental JailbreaksLLM safety and security → Jailbreaking
OpenMMReasoner: Pushing the Frontiers for Multimodal Reasoning with an Open and General RecipeLLM reasoning → Multimodal reasoning
Beyond Generative AI: World Models for Clinical Prediction, Counterfactuals, and PlanningWorld models → Medical and surgical world models
Incorporating Self-Rewriting into Large Language Model Reasoning ReinforcementLLM reasoning → Reinforcement learning for reasoning
Safe and Optimal Variable Impedance Control via Certified Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
ARK: Answer-Centric Retriever Tuning via KG-augmented Curriculum LearningRetrieval-augmented generation → RAG robustness and adaptive retrieval
SDA: Steering-Driven Distribution Alignment for Open LLMs without Fine-TuningLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
NaTex: Seamless Texture Generation as Latent Color DiffusionDiffusion models → 3D, point cloud and scene generation
Optimizing Operation Recipes with Reinforcement Learning for Safe and Interpretable Control of Chemical ProcessesDeep reinforcement learning methods → RL for energy, networks and operations
Spectral Identifiability for Interpretable Probe GeometryMechanistic interpretability → Circuits and mechanistic analysis
MuISQA: Multi-Intent Retrieval-Augmented Generation for Scientific Question AnsweringRetrieval-augmented generation → Agentic RAG, agents and memory
"To Survive, I Must Defect": Jailbreaking LLMs via the Game-Theory ScenariosLLM safety and security → Jailbreaking
FT-NCFM: An Influence-Aware Data Distillation Framework for Efficient VLA ModelsVision-language-action models → VLA architectures and generalization
Pass@k Metric for RLVR: A Diagnostic Tool of Exploration, But Not an ObjectiveLLM reasoning → Reinforcement learning for reasoning
Q-MLLM: Vector Quantization for Robust Multimodal Large Language Model SecurityMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
Deep SOR Minimax Q-learning for Two-player Zero-sum GameDeep reinforcement learning methods → Deep RL algorithms and representations
Can MLLMs Read the Room? A Multimodal Benchmark for Assessing Deception in Multi-Party Social InteractionsLLM reasoning → Prompting and chain-of-thought reasoning +1
Towards Overcoming Data Scarcity in Nuclear Energy: A Study on Critical Heat Flux with Physics-consistent Conditional Diffusion ModelDiffusion models → Flow matching, time series and science
ChemLabs on ChemO: A Multi-Agent System for Multimodal Reasoning on IChO 2025LLM reasoning → Clinical and biomedical applications
When Alignment Fails: Multimodal Adversarial Attacks on Vision-Language-Action ModelsVision-language-action models → Safety and adversarial attacks
Multi-Agent Collaborative Reward Design for Enhancing Reasoning in Reinforcement LearningLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
SemanticCite: Citation Verification with AI-Powered Full-Text Analysis and Evidence-Based ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning
Fast LLM Post-training via Decoupled and Fastest-of-N SpeculationEfficient LLM inference and serving → Speculative decoding
Mantis: A Versatile Vision-Language-Action Model with Disentangled Visual ForesightVision-language-action models → World model-based VLA
EvoVLA: Self-Evolving Vision-Language-Action ModelVision-language-action models → Reinforcement learning and post-training
An Image Is Worth Ten Thousand Words: Verbose-Text Induction Attacks on VLMsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
Video2Layout: Recall and Reconstruct Metric-Grounded Cognitive Map for Spatial ReasoningLLM reasoning → Spatial and embodied reasoning +1
MagBotSim: Physics-Based Simulation and Reinforcement Learning Environments for Magnetic RoboticsDeep reinforcement learning methods → RL for robot locomotion and manipulation
Pluggable Pruning with Contiguous Layer Distillation for Diffusion TransformersDiffusion models → Diffusion foundations and sampling
Reasoning Guided Embeddings: Leveraging MLLM Reasoning for Improved Multimodal RetrievalLLM reasoning → Multimodal reasoning
Real-Time 3D Object Detection with Inference-Aligned LearningKnowledge distillation → Detection and dense prediction distillation
Multidimensional Rubric-oriented Reward Model Learning via Geometric Projection Reference ConstraintsLLM alignment / RLHF / preference optimisation → Medical LLM alignment
An Interpretability-Guided Framework for Responsible Synthetic Data Generation in Emotional TextSynthetic data and self-improvement for LLMs → Privacy-preserving tabular data synthesis
AskDB: An LLM Agent for Natural Language Interaction with Relational DatabasesLLM agents and tool use → Web, search, and research agents
Decoupling Complexity from Scale in Latent Diffusion ModelDiffusion models → Diffusion foundations and sampling
SkyRL-Agent: Efficient RL Training for Multi-turn LLM AgentLLM agents and tool use → Reinforcement learning for LLM agents
Mitigating Estimation Bias with Representation Learning in TD Error-Driven RegularizationDeep reinforcement learning methods → RL theory and sample complexity
Mathematical Framework for Custom Reward Functions in Job Application Evaluation using Reinforcement LearningDeep reinforcement learning methods → RL training of language and agent models
Bellman Memory Units: A neuromorphic framework for synaptic reinforcement learning with an evolving network topologyDeep reinforcement learning methods → Spiking and neuromorphic RL
LiSTAR: Ray-Centric World Models for 4D LiDAR Sequences in Autonomous DrivingWorld models → World modeling from LiDAR generation +1
AMS-KV: Adaptive KV Caching in Multi-Scale Visual Autoregressive TransformersEfficient LLM inference and serving → Fast autoregressive visual generation
Agent0: Unleashing Self-Evolving Agents from Zero Data via Tool-Integrated ReasoningLLM reasoning → Reasoning for agents, planning and code +2
Mixture of Ranks with Degradation-Aware Routing for One-Step Real-World Image Super-ResolutionMixture-of-experts → Vision mixture-of-experts
CARE: Turning LLMs Into Causal Reasoning ExpertLLM reasoning → Causal reasoning in LLMs
Hiding in the AI Traffic: Abusing MCP for LLM-Powered Agentic Red TeamingLLM agents and tool use → Agent threats, guardrails and governance +1
Sensorium Arc: AI Agent System for Oceanic Data Exploration and Interactive Eco-ArtLLM agents and tool use → Multimodal and embodied agents
CARE-RAG - Clinical Assessment and Reasoning in RAGLLM reasoning → Clinical and biomedical applications
Detecting Sleeper Agents in Large Language Models via Semantic Drift AnalysisLLM agents and tool use → Agent threats, guardrails and governance +1
Fairness in Multi-modal Medical Diagnosis with Demonstration SelectionMultimodal LLMs / vision-language models → Medical vision-language models
Descend or Rewind? Stochastic Gradient Descent UnlearningLLM safety and security → Machine unlearning
KRAL: Knowledge and Reasoning Augmented Learning for LLM-assisted Clinical Antimicrobial TherapyLLM reasoning → Clinical and biomedical applications

All 348