Papers

New on arXiv

Mon 3 Nov 2025

343 papers

Democratizing LLM Efficiency: From Hyperscale Optimizations to Universal DeployabilityRetrieval-augmented generation → RAG pipeline design and evaluation +2
Temporal Predictors of Outcome in Reasoning Language ModelsLLM reasoning → Test-time compute and efficient thinking
Preference Orchestrator: Prompt-Aware Multi-Objective Alignment for Large Language ModelsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
AudAgent: Automated Auditing of Privacy Policy Compliance in AI AgentsLLM agents and tool use → Agent threats, guardrails and governance
Compressing Multi-Task Model for Autonomous Driving via Pruning and Knowledge DistillationKnowledge distillation → Detection and dense prediction distillation
EVLP:Learning Unified Embodied Vision-Language Planner with Reinforced Supervised Fine-TuningLLM reasoning → Spatial and embodied reasoning
Mathematical exploration and discovery at scaleLLM agents and tool use → Educational agents and math problem solving
Re-FORC: Adaptive Reward Prediction for Efficient Chain-of-Thought ReasoningLLM reasoning → Test-time compute and efficient thinking
InsurAgent: A Large Language Model-Empowered Agent for Simulating Individual Behavior in Purchasing Flood InsuranceLLM agents and tool use → Financial trading and forecasting agents
Enhancing Multimodal Recommendations with Vision-Language Models and Information-Aware FusionMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
Deep Value Benchmark: Measuring Whether Models Generalize Deep Values or Shallow PreferencesLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Second-Order Policy Gradient Methods for the Linear Quadratic RegulatorDeep reinforcement learning methods → RL for games and combinatorial optimization
Automated Reward Design for Gran TurismoLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Natural Building Blocks for Structured World Models: Theory, Evidence, and ScalingWorld models → Latent predictive world models
Watermarking Discrete Diffusion Language ModelsDiffusion models → Robustness, privacy and adversarial aspects +1
Beyond Static Cutoffs: One-Shot Dynamic Thresholding for Diffusion Language ModelsDiffusion models → Discrete and text diffusion
Human-AI Co-Embodied Intelligence for Scientific Experimentation and ManufacturingLLM agents and tool use → Multimodal and embodied agents
Text-VQA Aug: Pipelined Harnessing of Large Multimodal Models for Automated SynthesisMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Regularization Through Reasoning: Systematic Improvements in Language Model Classification via Explanation-Enhanced Fine-TuningLLM reasoning → Reasoning benchmarks and evaluation
TapOut: A Bandit-Based Approach to Dynamic Speculative DecodingEfficient LLM inference and serving → Speculative decoding
TRACE: Textual Reasoning for Affordance Coordinate ExtractionLLM reasoning → Spatial and embodied reasoning +1
Black-Box Membership Inference Attack for LVLMs via Prior Knowledge-Calibrated Memory ProbingMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
Tool-to-Agent Retrieval: Bridging Tools and Agents for Scalable LLM Multi-Agent SystemsLLM agents and tool use → Tool learning and function calling
Towards Robust Mathematical ReasoningLLM reasoning → Mathematical reasoning
Priors in Time: Missing Inductive Biases for Language Model InterpretabilityMechanistic interpretability → SAE features and their applications
TIR-Bench: A Comprehensive Benchmark for Agentic Thinking-with-Images ReasoningLLM reasoning → Multimodal reasoning +1
Routing-Based Continual Learning for Multimodal Large Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Simulating Environments with Reasoning Models for Agent TrainingLLM reasoning → Reasoning for agents, planning and code +1
KV Cache Transform Coding for Compact Storage in LLM InferenceEfficient LLM inference and serving → KV cache optimization
PROPEX-RAG: Enhanced GraphRAG using Prompt-Driven Prompt ExecutionRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
How Far Are Surgeons from Surgical World Models? A Pilot Study on Zero-shot Surgical Video Generation with Expert AssessmentWorld models → Medical and surgical world models
UniLION: Towards Unified Autonomous Driving Model with Linear Group RNNsEnd-to-end autonomous driving → End-to-end system architecture
Wonder3D++: Cross-domain Diffusion for High-fidelity 3D Generation from a Single ImageDiffusion models → 3D, point cloud and scene generation
RLAC: Reinforcement Learning with Adversarial Critic for Free-Form Generation TasksDeep reinforcement learning methods → RL training of language and agent models
Towards Efficient Federated Learning of Networked Mixture-of-Experts for Mobile Edge ComputingMixture-of-experts → Federated mixture-of-experts
Efficient Tool-Calling Multi-Expert NPC Agent for Commonsense Persona-Grounded DialogueLLM agents and tool use → Social simulation and multi-agent debate
Unified Diffusion VLA: Vision-Language-Action Model via Joint Discrete Denoising Diffusion ProcessVision-language-action models → Efficient VLA inference and adaptation +1
Collaborative Large Language Model Inference via Resource-Aware Parallel Speculative DecodingEfficient LLM inference and serving → Speculative decoding
Bayesian Natural Gradient Fine-Tuning of CLIP Models via Kalman FilteringMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Open Character Training: Shaping the Persona of AI Assistants through Constitutional AILLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
UniLumos: Fast and Unified Image and Video Relighting with Physics-Plausible FeedbackDiffusion models → 3D, point cloud and scene generation
Hybrid Retrieval-Augmented Generation Agent for Trustworthy Legal Question Answering in Judicial ForensicsRetrieval-augmented generation → Legal and financial RAG +1
EngTrace: A Symbolic Benchmark for Verifiable Process Supervision of Engineering ReasoningLLM reasoning → Reasoning benchmarks and evaluation
Evaluating Cultural Knowledge Processing in Large Language Models: A Cognitive Benchmarking Framework Integrating Retrieval-Augmented GenerationRetrieval-augmented generation → Generation and context integration
Enhancing Diffusion-based Restoration Models via Difficulty-Adaptive Reinforcement Learning with IQA RewardDiffusion models → Diffusion and flow policies for robotics +1
A Graph-based RAG for Energy Efficiency Question AnsweringRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
Prompt Injection as an Emerging Threat: Evaluating the Resilience of Large Language ModelsLLM safety and security → Prompt injection and agent security
Scaling Graph Chain-of-Thought Reasoning: A Multi-Agent Framework with Efficient LLM ServingLLM reasoning → Knowledge- and tool-augmented reasoning +2
Actial: Activate Spatial Reasoning Ability of Multimodal Large Language ModelsLLM reasoning → Spatial and embodied reasoning +1
MARS: Multi-Agent Robotic System with Multimodal Large Language Models for Assistive IntelligenceMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
PixelVLA: Advancing Pixel-level Understanding in Vision-Language-Action ModelVision-language-action models → Generalist VLA models and benchmarks
Learning what to say and how precisely: Efficient Communication via Differentiable Discrete Communication LearningDeep reinforcement learning methods → RL for energy, networks and operations
TPS-Bench: Evaluating AI Agents' Tool Planning & Scheduling Abilities in Compounding TasksLLM agents and tool use → General LLM agent frameworks and benchmarks
Synthetic Eggs in Many Baskets: The Impact of Synthetic Data Diversity on LLM Fine-TuningSynthetic data and self-improvement for LLMs → LLM synthetic data generation
AERMANI-VLM: Structured Prompting and Reasoning for Aerial Manipulation with Vision Language ModelsLLM reasoning → Spatial and embodied reasoning +1
BARD: budget-aware reasoning distillationLLM reasoning → Test-time compute and efficient thinking +1
HMVLM: Human Motion-Vision-Lanuage Model via MoE LoRAMixture-of-experts → MoE in reinforcement learning and robotics
Reg-DPO: SFT-Regularized Direct Preference Optimization with GT-Pair for Improving Video GenerationLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
LiCoMemory: Lightweight and Cognitive Agentic Memory for Efficient Long-Term ReasoningLLM reasoning → Reasoning for agents, planning and code +1
Towards One-step Causal Video Generation via Adversarial Self-DistillationKnowledge distillation → Multimodal distillation
Modulation of temporal decision-making in a deep reinforcement learning agent under the dual-task paradigmDeep reinforcement learning methods → Deep RL algorithms and representations
LiveSearchBench: An Automatically Constructed Benchmark for Retrieval and Reasoning over Dynamic KnowledgeLLM reasoning → Knowledge- and tool-augmented reasoning
RAGSmith: A Framework for Finding the Optimal Composition of Retrieval-Augmented Generation Methods Across DatasetsRetrieval-augmented generation → RAG pipeline design and evaluation
Protecting the Neural Networks against FGSM Attack Using Machine UnlearningLLM safety and security → Machine unlearning
Align to Misalign: Automatic LLM Jailbreak with Meta-Optimized LLM JudgesLLM safety and security → Jailbreaking
Learning Intractable Multimodal Policies with Reparameterization and Diversity RegularizationDeep reinforcement learning methods → RL with diffusion and flow models
The Ouroboros of Benchmarking: Reasoning Evaluation in an Era of SaturationLLM reasoning → Reasoning benchmarks and evaluation
CMI-MTL: Cross-Mamba interaction based multi-task learning for medical visual question answeringState-space sequence models (S4, Mamba) → Vision Mamba backbones
Expanding the Content-Style Frontier: a Balanced Subspace Blending Approach for Content-Style LoRA FusionDiffusion models → Image generation, text-to-image and editing
Thinking with DistilQwen: A Tale of Four Distilled Reasoning and Reward Model SeriesLLM reasoning → Prompting and chain-of-thought reasoning +2
left|,circlearrowright,boxedBUS,right|: A Large and Diverse Multimodal Benchmark for evaluating the ability of Vision-Language Models to understand Rebus PuzzlesLLM reasoning → Multimodal reasoning +1
Embodied Cognition Augmented End2End Autonomous DrivingEnd-to-end autonomous driving → End-to-end system architecture
RobustVLA: Robustness-Aware Reinforcement Post-Training for Vision-Language-Action ModelsVision-language-action models → Reinforcement learning and post-training
MVSMamba: Multi-View Stereo with State Space ModelState-space sequence models (S4, Mamba) → Vision Mamba backbones
From Pixels to Cooperation Multi Agent Reinforcement Learning based on Multimodal World ModelsWorld models → Latent dynamics models and model-based RL
Perturb a Model, Not an Image: Towards Robust Privacy Protection via Anti-Personalized Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects
FirstAidQA: A Synthetic Dataset for First Aid and Emergency Response in Low-Connectivity SettingsSynthetic data and self-improvement for LLMs → Medical synthetic data and agents
"Give a Positive Review Only": An Early Investigation Into In-Paper Prompt Injection Attacks and Defenses for AI ReviewersLLM safety and security → Prompt injection and agent security
When, What, and How: Rethinking Retrieval-Enhanced Speculative DecodingLLM reasoning → Test-time compute and efficient thinking +1
Contact Map Transfer with Conditional Diffusion Model for Generalizable Dexterous Grasp GenerationDiffusion models → Diffusion and flow policies for robotics
Rescuing the Unpoisoned: Efficient Defense against Knowledge Corruption Attacks on RAG SystemsRetrieval-augmented generation → Privacy and security of RAG
Don't Just Search, Understand: Semantic Path Planning Agent for Spherical Tensegrity Robots in Unknown EnvironmentsLLM agents and tool use → Multimodal and embodied agents
Embodiment Transfer Learning for Vision-Language-Action ModelsVision-language-action models → Human-video and latent-action pretraining
Saliency-Guided Domain Adaptation for Left-Hand Driving in Autonomous SteeringEnd-to-end autonomous driving → End-to-end system architecture
Optimizing Electric Vehicle Charging Station Placement Using Reinforcement Learning and Agent-Based SimulationsDeep reinforcement learning methods → RL for energy, networks and operations
Thought-For-Food: Reasoning Chain Induced Food Visual Question AnsweringLLM reasoning → Multimodal reasoning
OmniVLA: Physically-Grounded Multimodal VLA with Unified Multi-Sensor Perception for Robotic ManipulationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
FEval-TTC: Fair Evaluation Protocol for Test-Time ComputeLLM reasoning → Test-time compute and efficient thinking
DEER: Disentangled Mixture of Experts with Instance-Adaptive Routing for Generalizable Machine-Generated Text DetectionMixture-of-experts → Expert routing strategies
Self-Harmony: Learning to Harmonize Self-Supervision and Self-Play in Test-Time Reinforcement LearningDeep reinforcement learning methods → RL training of language and agent models
Analyzing the Power of Chain of Thought through Memorization CapabilitiesLLM reasoning → Prompting and chain-of-thought reasoning
QiMeng-NeuComBack: Self-Evolving Translation from IR to Assembly CodeSynthetic data and self-improvement for LLMs → Code and instruction data synthesis
MiRAGE: Misconception Detection with Retrieval-Guided Multi-Stage Reasoning and Ensemble FusionLLM reasoning → Mathematical reasoning
Scaling Cross-Embodiment World Models for Dexterous ManipulationWorld models → Embodied and robot world models
Diffusion Transformer meets Multi-level Wavelet Spectrum for Single Image Super-ResolutionDiffusion models → Inverse problems and image restoration
Conditional Diffusion Model-Enabled Scenario-Specific Neural Receivers for Superimposed Pilot SchemesDiffusion models → Diffusion for compression and wireless
DART: Difficulty-Adaptive Reasoning Truncation for Efficient Large Language ModelsLLM reasoning → Test-time compute and efficient thinking
MicroRemed: Benchmarking LLMs in Microservices RemediationLLM reasoning → Reasoning for agents, planning and code
Modular Task Decomposition and Dynamic Collaboration in Multi-Agent Systems Driven by Large Language ModelsLLM agents and tool use → General LLM agent frameworks and benchmarks
Uncovering EEG Patterns Consistently Associated with Cybersickness Discomfort Using Deep Learning Interpretability Maps

All 343