Papers

New on arXiv

Mon 5 Jan 2026

313 papers

Evaluating Reward Model Generalization via Pairwise Maximum Discrepancy CompetitionsLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Crystal-KV: Efficient KV Cache Management for Chain-of-Thought LLMs via Answer-First PrincipleLLM reasoning → Test-time compute and efficient thinking +1
Domain Specific Specialization in Low-Resource Settings: The Efficacy of Offline Response-Based Knowledge Distillation in Large Language ModelsKnowledge distillation → Knowledge distillation for LLMs
EPD-Serve: A Flexible Multimodal EPD Disaggregation Inference Serving System On AscendMultimodal LLMs / vision-language models → Visual token pruning and efficiency
OrbitFlow: SLO-Aware Long-Context LLM Serving with Fine-Grained KV Cache ReconfigurationEfficient LLM inference and serving → Serving systems and request scheduling
An LLM -Powered Assessment Retrieval-Augmented Generation (RAG) For Higher EducationRetrieval-augmented generation → Agentic RAG, agents and memory
A Review of Online Diffusion Policy RL Algorithms for Scalable Robotic ControlDiffusion models → Diffusion and flow policies for robotics +1
AgentMark: Utility-Preserving Behavioral Watermarking for AgentsLLM agents and tool use → Agent threats, guardrails and governance +1
Scalable Construction of a Lung Cancer Knowledge Base: Profiling Semantic Reasoning in LLMsLLM reasoning → Clinical and biomedical applications
LongDA: Benchmarking LLM Agents for Long-Document Data AnalysisLLM agents and tool use → Web, search, and research agents
Annealed Langevin Posterior Sampling (ALPS): A Rapid Algorithm for Image Restoration with Multiscale Energy ModelsDiffusion models → Inverse problems and image restoration
Orchestral AI: A Framework for Agent OrchestrationLLM agents and tool use → Software engineering coding agents
LoRA-Drop: Temporal LoRA Decoding for Efficient LLM InferenceEfficient LLM inference and serving → Early exit and edge inference
Shallow- and Deep-fake Image Manipulation Localization Using Vision Mamba and Guided Graph Neural NetworkState-space sequence models (S4, Mamba) → Vision Mamba backbones
SimpleMem: Efficient Lifelong Memory for LLM AgentsLLM agents and tool use → Agent memory systems
On the Effectiveness of Proposed Techniques to Reduce Energy Consumption in RAG Systems: A Controlled ExperimentRetrieval-augmented generation → RAG pipeline design and evaluation
Textual Explanations and Their Evaluations for Reinforcement Learning PolicyDeep reinforcement learning methods → Interpretable and logic-guided RL
Polynomial Convergence of Riemannian Diffusion ModelsDiffusion models → Diffusion foundations and sampling
InternVLA-A1: Unifying Understanding, Generation and Action for Robotic ManipulationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
The Rise of Agentic Testing: Multi-Agent Systems for Robust Software Quality AssuranceLLM agents and tool use → Software engineering coding agents
VocalBridge: Latent Diffusion-Bridge Purification for Defeating Perturbation-Based Voiceprint DefensesDiffusion models → Audio, speech and music
Evaluating the Diagnostic Classification Ability of Multimodal Large Language Models: Insights from the Osteoarthritis InitiativeMultimodal LLMs / vision-language models → Medical vision-language models
Understanding Pure Textual Reasoning for Blind Image Quality AssessmentLLM reasoning → Multimodal reasoning
WebGym: Scaling Training Environments for Visual Web Agents with Realistic TasksLLM agents and tool use → Web, search, and research agents
ExposeAnyone: Personalized Audio-to-Expression Diffusion Models Are Robust Zero-Shot Face Forgery DetectorsDiffusion models → Image generation, text-to-image and editing
Talk2Move: Reinforcement Learning for Text-Instructed Object-Level Geometric Transformation in ScenesDeep reinforcement learning methods → RL with diffusion and flow models
Falcon-H1R: Pushing the Reasoning Frontiers with a Hybrid Model for Efficient Test-Time ScalingLLM reasoning → Test-time compute and efficient thinking
Project Ariadne: A Structural Causal Framework for Auditing Faithfulness in LLM AgentsLLM reasoning → Prompting and chain-of-thought reasoning +1
CycleVLA: Proactive Self-Correcting Vision-Language-Action Models via Subtask Backtracking and Minimum Bayes Risk DecodingVision-language-action models → Failure recovery and robustness
CD4LM: Consistency Distillation and aDaptive Decoding for Diffusion Language ModelsDiffusion models → Discrete and text diffusion
FMVP: Masked Flow Matching for Adversarial Video PurificationDiffusion models → Robustness, privacy and adversarial aspects
From XAI to Stories: A Factorial Study of LLM-Generated Explanation QualityLLM reasoning → Reasoning benchmarks and evaluation
TexTailor: Inference-Time Textual Guidance Tailoring for Multimodal Diffusion TransformersDiffusion models → Image generation, text-to-image and editing +1
CORE: Code-based Inverse Self-Training Framework with Graph Expansion for Virtual AgentsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +2
Code for Machines, Not Just Humans: Quantifying AI-Friendliness with Code Health MetricsLLM agents and tool use → Software engineering coding agents
Streaming Hallucination Detection in Long Chain-of-Thought ReasoningLLM reasoning → Hallucination detection and mitigation
EverMemOS: A Self-Organizing Memory Operating System for Structured Long-Horizon ReasoningLLM reasoning → Reasoning for agents, planning and code
Multi-fidelity graph-based neural networks architectures to learn Navier-Stokes solutions on non-parametrized 2D domainsState-space sequence models (S4, Mamba) → Time series forecasting with SSMs
BiPrompt: Bilateral Prompt Optimization for Visual and Textual Debiasing in Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Routing by Analogy: kNN-Augmented Expert Assignment for Mixture-of-ExpertsMixture-of-experts → Expert routing strategies
A Mamba-Based Model for Automatic Chord RecognitionState-space sequence models (S4, Mamba) → Speech and audio SSMs
Realistic adversarial scenario generation via human-like pedestrian model for autonomous vehicle control parameter optimisationEnd-to-end autonomous driving → Adversarial scenarios and safety testing
Deferred Commitment Decoding for Diffusion Language ModelsDiffusion models → Discrete and text diffusion
Cost-Efficient Cross-Lingual Retrieval-Augmented Generation for Low-Resource Languages: A Case Study in Bengali Agricultural AdvisoryRetrieval-augmented generation → Time series and agricultural RAG
Higher-Order Action Regularization in Deep Reinforcement Learning: From Continuous Control to Building Energy ManagementDeep reinforcement learning methods → RL for energy, networks and operations
Agentic Retoucher for Text-To-Image GenerationDiffusion models → Image generation, text-to-image and editing +1
GDRO: Group-level Reward Post-training Suitable for Diffusion ModelsDiffusion models → Diffusion and flow policies for robotics +1
Not All Needles Are Found: How Fact Distribution and Don't Make It Up Prompts Shape Retrieval, Reasoning, and Hallucination in Long-Context LLMsLLM reasoning → Hallucination detection and mitigation
Enhancing Object Detection with Privileged Information: A Model-Agnostic Teacher-Student ApproachKnowledge distillation → Detection and dense prediction distillation
Surprisal and Metaphor Novelty Judgments: Moderate Correlations and Divergent Scaling Effects Revealed by Corpus-Based and Synthetic DatasetsSynthetic data and self-improvement for LLMs → LLM synthetic data generation
XAI-MeD: Explainable Knowledge Guided Neuro-Symbolic Framework for Domain Generalization and Rare Class Detection in Medical ImagingLLM reasoning → Clinical and biomedical applications
Exploring Approaches for Detecting Memorization of Recommender System Data in Large Language ModelsLLM safety and security → Machine unlearning
Thinking with Blueprints: Assisting Vision-Language Models in Spatial Reasoning via Structured Object RepresentationLLM reasoning → Spatial and embodied reasoning +1
ChaosBench-Logic: A Benchmark for Logical and Symbolic Reasoning on Chaotic Dynamical SystemsLLM reasoning → Reasoning benchmarks and evaluation
Hidden State Poisoning Attacks against Mamba-based Language ModelsState-space sequence models (S4, Mamba) → Core SSM architectures for language
Forget Less by Learning Together through Concept ConsolidationDiffusion models → Robustness, privacy and adversarial aspects
AFTER: Mitigating the Object Hallucination of LVLM via Adaptive Factual-Guided Activation EditingMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs +1
MotionAdapter: Video Motion Transfer via Content-Aware Attention CustomizationDiffusion models → Video generation
Learning Diffusion Policy from Primitive Skills for Robot ManipulationDiffusion models → Diffusion and flow policies for robotics
D'ej`aQ: Open-Ended Evolution of Diverse, Learnable and Verifiable ProblemsLLM reasoning → Mathematical reasoning
Distorted Distributional Policy Evaluation for Offline Reinforcement LearningOffline RL and imitation learning from data → Offline RL algorithms
TalkPhoto: A Versatile Training-Free Conversational Assistant for Intelligent Image EditingMultimodal LLMs / vision-language models → Image editing and quality assessment
Evaluating Feature Dependent Noise in Preference-based Reinforcement LearningLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF +1
FedBiCross: Personalized One-Shot Federated Learning on Medical ImagesKnowledge distillation → Federated, robust and private distillation
Tackling the Inherent Difficulty of Noise Filtering in RAGRetrieval-augmented generation → RAG robustness and adaptive retrieval
Forget Less by Learning from Parents Through Hierarchical RelationshipsDiffusion models → Robustness, privacy and adversarial aspects
Agentic AI in Remote Sensing: Foundations, Taxonomy, and Emerging SystemsLLM agents and tool use → Multimodal and embodied agents
Safety at One Shot: Patching Fine-Tuned LLMs with A Single InstanceLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks
Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model AgentsLLM agents and tool use → Agent memory systems
Toward Auditable Neuro-Symbolic Reasoning in Pathology: SQL as an Explicit Trace of EvidenceLLM reasoning → Clinical and biomedical applications
CogFlow: Bridging Perception and Reasoning through Knowledge Internalization for Visual Mathematical Problem SolvingLLM reasoning → Multimodal reasoning +1
DermoGPT: Open Weights and Open Data for Morphology-Grounded Dermatological Reasoning MLLMsLLM reasoning → Clinical and biomedical applications +1
Jenius Agent: Towards Experience-Driven Accuracy Optimization in Real-World ScenariosLLM agents and tool use → Tool learning and function calling
Clinical Knowledge Graph Construction and Evaluation with Multi-LLMs via Retrieval-Augmented GenerationRetrieval-augmented generation → Medical and biomedical RAG
CSCBench: A PVC Diagnostic Benchmark for Commodity Supply Chain ReasoningLLM reasoning → Reasoning for agents, planning and code
V-CORE: Temporally Consistent Video Understanding for Video-LLMLLM reasoning → Multimodal reasoning
Moments Matter:Stabilizing Policy Optimization using Return DistributionsDeep reinforcement learning methods → RL theory and sample complexity
Sparse Threats, Focused Defense: Criticality-Aware Robust Reinforcement Learning for Safe Autonomous DrivingEnd-to-end autonomous driving → Adversarial scenarios and safety testing +1
VerLM: Explaining Face Verification Using Natural LanguageMultimodal LLMs / vision-language models → Social bias and fairness in VLMs
UnPII: Unlearning Personally Identifiable Information with Quantifiable Exposure RiskLLM safety and security → Machine unlearning
SRAS: A Lightweight Reinforcement Learning-based Document Selector for Edge-Native RAG PipelinesRetrieval-augmented generation → Agentic RAG, agents and memory +1
AlignDrive: Aligned Lateral-Longitudinal Planning for End-to-End Autonomous DrivingEnd-to-end autonomous driving → Imitation-learning end-to-end planners
Crafting Adversarial Inputs for Large Vision-Language Models Using Black-Box OptimizationMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
AI Agent Systems: Architectures, Applications, and EvaluationLLM agents and tool use → General LLM agent frameworks and benchmarks
Local Layer-wise Differential Privacy in Federated LearningLLM safety and security → Federated unlearning and poisoning
Yuan3.0 Flash: An Open Multimodal Large Language Model for Enterprise ApplicationsLLM reasoning → Test-time compute and efficient thinking +2
Entropy-Aligned Decoding of LMs for Better Writing and ReasoningLLM reasoning → Test-time compute and efficient thinking
Reinforcement Learning for Option Hedging: Static Implied-Volatility Fit versus Shortfall-Aware PerformanceDeep reinforcement learning methods → Adversarial robustness and RL in finance
A Training-Free Large Reasoning Model-based Knowledge Tracing Framework for Unified Prediction and PrescriptionLLM reasoning → Test-time compute and efficient thinking
Explicit World Models for Reliable Human-Robot CollaborationWorld models → World models for navigation
Varying-coefficient mixture of experts model for dynamic heterogeneous populations: application to mouse cortical developmentMixture-of-experts → Medical and biomedical MoE
ELF: A Family of Encoder-Free ECG-Language Models
A Foundation Model Approach for Fetal Stress Prediction During Labor From cardiotocography (CTG) recordings
LLM Flow Processes for Text-Conditioned Regression
Bridging the AI divide in sub-Saharan Africa: Challenges and opportunities for inclusivity
The Patient/Industry Trade-off in Medical Artificial Intelligence
Emergent Complexity in Nuclear Reaction Networks: A Study of Stellar Nucleosynthesis through Chemical Organization Theory
Is Sanskrit the most token-efficient language? A quantitative study using GPT, Gemini, and SentencePiece
Causal and Federated Multimodal Learning for Cardiovascular Risk Prediction under Heterogeneous Populations
Low-Back Pain Physical Rehabilitation by Movement Analysis in Clinical Trial

All 313