Papers

New on arXiv

Thu 8 Jan 2026

400 papers

Web Retrieval-Aware Chunking (W-RAC) for Efficient and Cost-Effective Retrieval-Augmented Generation SystemsRetrieval-augmented generation → Multimodal RAG
ALL-FEM: Agentic Large Language models Fine-tuned for Finite Element MethodsLLM agents and tool use → Multimodal and embodied agents
ASKD-Whisper: Adaptive Self-knowledge Distillation for Efficient and Low-Latency Automatic Speech RecognitionKnowledge distillation → Speech and audio distillation
Hindsight Preference Replay Improves Preference-Conditioned Multi-Objective Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
MLLM-VADStory: Domain Knowledge-Driven Multimodal LLMs for Video Ad Storyline InsightsMultimodal LLMs / vision-language models → Video understanding with MLLMs
SPINAL -- Scaling-law and Preference Integration in Neural Alignment LayersLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
LLM Agents in Law: Taxonomy, Applications, and ChallengesLLM agents and tool use → Social simulation and multi-agent debate
Akasha 2: Hamiltonian State Space Duality and Visual-Language Joint Embedding Predictive ArchitecturDiffusion models → Image generation, text-to-image and editing +1
Cascading multi-agent anomaly detection in surveillance systems via vision-language models and embedding-based classificationMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
QwenStyle: Content-Preserving Style Transfer with Qwen-Image-EditDiffusion models → Image generation, text-to-image and editing
Leveraging Membership Inference Attacks for Privacy Measurement in Federated Learning for Remote Sensing ImagesLLM safety and security → Federated unlearning and poisoning
FastSLM: Hierarchical Temporal Abstraction for Efficient Long-Form Speech AdaptationMultimodal LLMs / vision-language models → Visual token pruning and efficiency
Rational Synthesizers or Heuristic Followers? Analyzing LLMs in RAG-based Question-AnsweringRetrieval-augmented generation → RAG robustness and adaptive retrieval
Thinking with Map: Reinforced Parallel Map-Augmented Agent for GeolocalizationLLM reasoning → Spatial and embodied reasoning +1
Interactive Distillation for Cooperative Multi-Agent Reinforcement LearningKnowledge distillation → On-policy and self-distillation for LLMs
Conformity and Social Impact on AI AgentsLLM agents and tool use → Social simulation and multi-agent debate
Imitation Learning for Combinatorial Optimisation under UncertaintyOffline RL and imitation learning from data → Imitation learning methods
Lost in Execution: On the Multilingual Robustness of Tool Calling in Large Language ModelsLLM agents and tool use → Tool learning and function calling
GlyRAG: Context-Aware Retrieval-Augmented Framework for Blood Glucose ForecastingRetrieval-augmented generation → Medical and biomedical RAG
Coding the Visual World: From Image to Simulation Using Vision Language ModelsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Multi-turn Jailbreaking Attack in Multi-Modal Large Language ModelsLLM safety and security → Vision-language and text-to-image safety
Ontology Neural Networks for Topologically Conditioned Constraint SatisfactionLLM reasoning → Reasoning benchmarks and evaluation
Effects of personality steering on cooperative behavior in Large Language Model agentsLLM agents and tool use → Social simulation and multi-agent debate
TIME: Temporally Intelligent Meta-reasoning Engine for Context-Triggered Explicit ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning
MoEBlaze: Breaking the Memory Wall for Efficient MoE Training on Modern GPUsMixture-of-experts → Efficient MoE training and inference systems
A Survey of Agentic AI and Cybersecurity: Challenges, Opportunities and Use-case PrototypesLLM agents and tool use → Agent threats, guardrails and governance
RL-AWB: Deep Reinforcement Learning for Auto White Balance Correction in Low-Light Night-time ScenesDeep reinforcement learning methods → RL with diffusion and flow models
LaST0: Latent Spatio-Temporal Chain-of-Thought for Robotic Vision-Language-Action ModelLLM reasoning → Spatial and embodied reasoning +1
GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL OptimizationLLM reasoning → Reinforcement learning for reasoning +1
RoboVIP: Multi-View Video Generation with Visual Identity Prompting Augments Robot ManipulationDiffusion models → Diffusion and flow policies for robotics
Robust Reasoning as a Symmetry-Protected Topological PhaseLLM reasoning → Reasoning benchmarks and evaluation
Learning Latent Action World Models In The WildWorld models → Embodied and robot world models
MineNPC-Task: Task Suite for Memory-Aware Minecraft AgentsLLM agents and tool use → Reinforcement learning for LLM agents
Internal Representations as Indicators of Hallucinations in Agent Tool SelectionLLM agents and tool use → Medical and scientific domain agents
FlowLet: Conditional 3D Brain MRI Synthesis using Wavelet Flow MatchingDiffusion models → Medical imaging and segmentation
MoE3D: A Mixture-of-Experts Module for 3D ReconstructionMixture-of-experts → Vision mixture-of-experts
EARL: Energy-Aware Optimization of Liquid State Machines for Pervasive AIDeep reinforcement learning methods → RL for energy, networks and operations
Mechanisms of Prompt-Induced Hallucination in Vision-Language ModelsMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
AgentCompress: Task-Aware Compression for Affordable Large Language Model AgentsLLM agents and tool use → Agent memory systems
VideoAuto-R1: Video Auto Reasoning via Thinking Once, Answering TwiceLLM reasoning → Multimodal reasoning +1
FaST: Efficient and Effective Long-Horizon Forecasting for Large-Scale Spatial-Temporal Graphs via Mixture-of-ExpertsMixture-of-experts → MoE for time series and graphs
CoV: Chain-of-View Prompting for Spatial ReasoningLLM reasoning → Spatial and embodied reasoning
RelayLLM: Efficient Reasoning via Collaborative DecodingLLM reasoning → Test-time compute and efficient thinking
DocDancer: Towards Agentic Document-Grounded Information SeekingLLM agents and tool use → Web, search, and research agents
Vision-Language Introspection: Mitigating Overconfident Hallucinations in MLLMs via Interpretable Bi-Causal SteeringMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Safe Continual Reinforcement Learning Methods for Nonstationary Environments. Towards a Survey of the State of the ArtDeep reinforcement learning methods → Safe and constrained RL
PC2: Politically Controversial Content Generation via Jailbreaking Attacks on GPT-based Text-to-Image ModelsLLM safety and security → Vision-language and text-to-image safety
Multi-Scale Local Speculative Decoding for Image GenerationEfficient LLM inference and serving → Fast autoregressive visual generation
Distilling the Thought, Watermarking the Answer: A Principle Semantic Guided Watermark for Large Reasoning ModelsLLM reasoning → Prompting and chain-of-thought reasoning +1
VerseCrafter: Dynamic Realistic Video World Model with 4D Geometric ControlWorld models → Video-generation world models and simulators
Sequential Subspace Noise Injection Prevents Accuracy Collapse in Certified UnlearningLLM safety and security → Machine unlearning
LooseRoPE: Content-aware Attention Manipulation for Semantic HarmonizationDiffusion models → Image generation, text-to-image and editing
VERSE: Visual Embedding Reduction and Space Exploration. Clustering-Guided Insights for Training Data Enhancement in Visually-Rich Document UnderstandingMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
Agent-as-a-JudgeLLM agents and tool use → General LLM agent frameworks and benchmarks
GlimpRouter: Efficient Collaborative Inference by Glimpsing One Token of ThoughtsLLM reasoning → Test-time compute and efficient thinking
Nalar: An agent serving frameworkLLM agents and tool use → Software engineering coding agents
Controllable Memory Usage: Balancing Anchoring and Innovation in Long-Term Human-Agent InteractionLLM agents and tool use → Agent memory systems
Arabic Prompts with English Tools: A BenchmarkLLM agents and tool use → General LLM agent frameworks and benchmarks
Driving on RegistersEnd-to-end autonomous driving → End-to-end system architecture
Chain-of-Sanitized-Thoughts: Plugging PII Leakage in CoT of Large Reasoning ModelsLLM reasoning → Reasoning for agents, planning and code
SemPA: Improving Sentence Embeddings of Large Language Models through Semantic Preference AlignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Milestones over Outcome: Unlocking Geometric Reasoning with Sub-Goal Verifiable RewardLLM reasoning → Spatial and embodied reasoning +1
From Understanding to Engagement: Personalized pharmacy Video Clips via Vision Language Models (VLMs)Multimodal LLMs / vision-language models → Medical vision-language models
Reinforced Efficient Reasoning via Semantically Diverse ExplorationLLM reasoning → Reinforcement learning for reasoning
DeepWeightFlow: Re-Basined Flow Matching for Generating Neural Network WeightsDiffusion models → Flow matching, time series and science
Large language models can effectively convince people to believe conspiraciesLLM safety and security → Safety alignment, refusal and guardrails
How to Set the Learning Rate for Large-Scale Pre-training?Mixture-of-experts → MoE theory and gating functions
FinDeepForecast: A Live Multi-Agent System for Benchmarking Deep Research Agents in Financial ForecastingLLM agents and tool use → Financial trading and forecasting agents
ArcAligner: Adaptive Recursive Aligner for Compressed Context Embeddings in RAGRetrieval-augmented generation → Generation and context integration +1
OptiSet: Unified Optimizing Set Selection and Ranking for Retrieval-Augmented GenerationRetrieval-augmented generation → Generation and context integration
H'an D=an Xu'e B`u (Mimicry) or Q=ing Ch=u Y'u L'an (Mastery)? A Cognitive Perspective on Reasoning Distillation in Large Language ModelsLLM reasoning → Prompting and chain-of-thought reasoning +1
On the Hidden Objective Biases of Group-based Reinforcement LearningLLM reasoning → Reinforcement learning for reasoning +1
AlgBench: To What Extent Do Large Reasoning Models Understand Algorithms?LLM reasoning → Test-time compute and efficient thinking
Learning from Mistakes: Negative Reasoning Samples Enhance Out-of-Domain GeneralizationLLM reasoning → Test-time compute and efficient thinking
ConMax: Confidence-Maximizing Compression for Efficient Chain-of-Thought ReasoningLLM reasoning → Test-time compute and efficient thinking
A Unified Spoken Language Model with Injected Emotional-Attribution Thinking for Human-like InteractionLLM reasoning → Audio-language model reasoning
Safe Reinforcement Learning Beyond Baseline Control: A Hierarchical Framework for Space Triangle Tethered Formation SystemDeep reinforcement learning methods → RL for robot navigation and UAV control
T-Retriever: Tree-based Hierarchical Retrieval Augmented Generation for Textual GraphsRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
V-FAT: Benchmarking Visual Fidelity Against Text-biasLLM reasoning → Multimodal reasoning +1
Scaling Vision Language Models for Pharmaceutical Long Form Video Reasoning on Industrial GenAI PlatformLLM reasoning → Multimodal reasoning +1
SmartSearch: Process Reward-Guided Query Refinement for Search AgentsLLM agents and tool use → Web, search, and research agents
Flexible Manufacturing Systems Intralogistics: Dynamic Optimization of AGVs and Tool Sharing Using Coloured-Timed Petri Nets and Actor-Critic RL with Actions MaskingDeep reinforcement learning methods → RL for energy, networks and operations
Analyzing Message-Code Inconsistency in AI Coding Agent-Authored Pull RequestsLLM agents and tool use → Software engineering coding agents
Mind2Report: Expert-Level Commercial Report Synthesis via Cognitive Deep Research AgentLLM agents and tool use → Web, search, and research agents
Higher-Order Knowledge Representations for Agentic Scientific ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning +1
EvolSQL: Structure-Aware Evolution for Scalable Text-to-SQL Data SynthesisSynthetic data and self-improvement for LLMs → Synthetic reasoning data and bootstrapping
Orchestrating Intelligence: Confidence-Aware Routing for Efficient Multi-Agent Collaboration across Multi-Scale ModelsLLM agents and tool use → General LLM agent frameworks and benchmarks
A Navigational Approach for Comprehensive RAG via Traversal over Proposition GraphsRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
RAAR: Retrieval Augmented Agentic Reasoning for Cross-Domain Misinformation DetectionLLM reasoning → Reasoning for agents, planning and code +2
SOVABench: A Vehicle Surveillance Action Retrieval Benchmark for Multimodal Large Language ModelsMultimodal LLMs / vision-language models → Video understanding with MLLMs
DR-LoRA: Dynamic Rank LoRA for Fine-Tuning Mixture-of-Experts ModelsMixture-of-experts → Mixture of LoRA experts
AECV-Bench: Benchmarking Multimodal Models on Architectural and Engineering Drawings UnderstandingMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
SCALER:Synthetic Scalable Adaptive Learning Environment for ReasoningLLM reasoning → Reinforcement learning for reasoning
Thinking-Based Non-Thinking: Solving the Reward Hacking Problem in Training Hybrid Reasoning Models via Reinforcement LearningLLM reasoning → Reinforcement learning for reasoning +1
Defense Against Indirect Prompt Injection via Tool Result ParsingLLM agents and tool use → Agent threats, guardrails and governance +1
APEX: Academic Poster Editing Agentic ExpertLLM agents and tool use → Multimodal and embodied agents
Measurement-Consistent Langevin Corrector for Stabilizing Latent Diffusion Inverse Problem SolversDiffusion models → Inverse problems and image restoration
Belief in Authority: Impact of Authority in Multi-Agent Evaluation FrameworkLLM agents and tool use → Social simulation and multi-agent debate
AgentOCR: Reimagining Agent History via Optical Self-CompressionLLM agents and tool use → Agent memory systems
GeM-VG: Towards Generalized Multi-image Visual Grounding with Multimodal Large Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning

All 400