Papers

New on arXiv

Mon 15 Dec 2025

356 papers

TWICE: Modeling the Temporal Evolution of Personalized User Behavior via Event-Driven AgentsLLM agents and tool use → Agent memory systems
Adaptive GPU Resource Allocation for Multi-Agent Collaborative Reasoning in Serverless EnvironmentsLLM reasoning → Reasoning for agents, planning and code +1
Bilevel Optimization for Covert Memory Tampering in Heterogeneous Multi-Agent Architectures (XAMT)LLM agents and tool use → Agent threats, guardrails and governance
Toward Agentic Environments: GenAI and the Convergence of AI, Sustainability, and Human-Centric SpacesLLM agents and tool use → General LLM agent frameworks and benchmarks
Beyond Training: Enabling Self-Evolution of Agents with MOBIMEMLLM agents and tool use → Agent memory systems
Workflows vs Agents for Code TranslationLLM agents and tool use → Software engineering coding agents
SocialNav-MoE: A Mixture-of-Experts Vision Language Model for Socially Compliant Navigation with Reinforcement Fine-TuningMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
AOI: Context-Aware Multi-Agent Operations via Dynamic Scheduling and Hierarchical Memory CompressionLLM agents and tool use → Software engineering coding agents
Pattern-Guided Diffusion ModelsDiffusion models → Flow matching, time series and science
Hierarchical Multi-agent Large Language Model Reasoning for Autonomous Functional Materials DiscoveryLLM reasoning → Clinical and biomedical applications +1
Let's (not) just put things in Context: Test-Time Training for Long-Context LLMsLLM reasoning → Test-time compute and efficient thinking
Group-Theoretic Reinforcement Learning of Dynamical Decoupling SequencesDeep reinforcement learning methods → Quantum reinforcement learning
Coarse-to-Fine Hierarchical Alignment for UAV-based Human Detection using Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
Verification-Guided Context Optimization for Tool Calling via Hierarchical LLMs-as-EditorsLLM agents and tool use → Tool learning and function calling
Improvise, Adapt, Overcome -- Telescopic Adapters for Efficient Fine-tuning of Vision Language Models in Medical ImagingMultimodal LLMs / vision-language models → Medical vision-language models
Explainable reinforcement learning from human feedback to improve alignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods +1
The Double Life of Code World Models: Provably Unmasking Malicious Behavior Through Execution TracesWorld models → World models for language agents +1
Constrained Policy Optimization via Sampling-Based Weight-Space ProjectionDeep reinforcement learning methods → Safe and constrained RL
Semantic Grounding Index: Geometric Bounds on Context Engagement in RAG SystemsRetrieval-augmented generation → Generation and context integration
State-Dependent Refusal and Learned Incapacity in RLHF-Aligned Language ModelsLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks
Improving the Plausibility of Pressure Distributions Synthesized from Depth Image through Generative ModelingDiffusion models → Medical imaging and segmentation
STAR: STacked AutoRegressive Scheme for Unified Multimodal LearningMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Why Text Prevails: Vision May Undermine Multimodal Medical Decision MakingMultimodal LLMs / vision-language models → Medical vision-language models
DiffusionBrowser: Interactive Diffusion Previews via Multi-Branch DecodersDiffusion models → Video generation
AgentIAD: Agentic Industrial Anomaly Detection via Adaptive Memory AugmentationLLM agents and tool use → Software engineering coding agents
A Scientific Reasoning Model for Organic Synthesis Procedure GenerationLLM reasoning → Clinical and biomedical applications
Grab-3D: Detecting AI-Generated Videos from 3D Geometric Temporal ConsistencyDiffusion models → 3D, point cloud and scene generation
Towards Spatial Trace with Reasoning in Vision-Language Models for RoboticsLLM reasoning → Spatial and embodied reasoning +1
Comparative Analysis of LLM Abliteration Methods: A Cross-Architecture EvaluationLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
World Models for Learning Dexterous Hand-Object Interactions from Human VideosWorld models → Embodied and robot world models
MindDrive: A Vision-Language-Action Model for Autonomous Driving via Online Reinforcement LearningVision-language-action models → VLA for autonomous driving +3
StutterFuse: Mitigating Modality Collapse in Stuttering Detection with Jaccard-Weighted Metric Learning and Gated FusionRetrieval-augmented generation → Hallucination detection in RAG
Do-Undo Bench: Reversibility for Action Understanding in Image GenerationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Nemotron-Cascade: Scaling Cascaded Reinforcement Learning for General-Purpose Reasoning ModelsLLM reasoning → Reinforcement learning for reasoning
LongVie 2: Multimodal Controllable Ultra-Long Video World ModelWorld models → Video-generation world models and simulators
Lighting in Motion: Spatiotemporal HDR Lighting EstimationDiffusion models → 3D, point cloud and scene generation
Image Diffusion Preview with Consistency SolverDiffusion models → Diffusion foundations and sampling
ReFusion: A Diffusion Large Language Model with Parallel Autoregressive DecodingDiffusion models → Discrete and text diffusion +1
MMhops-R1: Multimodal Multi-hop ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning
Superposition as Lossy Compression: Measure with Sparse Autoencoders and Connect to Adversarial VulnerabilityMechanistic interpretability → Sparse autoencoder training and evaluation
Memory in the Age of AI AgentsLLM agents and tool use → Agent memory systems
Async Control: Stress-testing Asynchronous Control Measures for LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance
Janus: Disaggregating Attention and Experts for Scalable MoE InferenceMixture-of-experts → Efficient MoE training and inference systems
A Deep Learning Model of Mental Rotation Informed by Interactive VR ExperimentsWorld models → World models for language agents
Reinforcement Learning based 6-DoF Maneuvers for Microgravity Intravehicular Docking: A Simulation Study with Int-Ball2 in ISS-JEMDeep reinforcement learning methods → RL for robot navigation and UAV control
MedCEG: Reinforcing Verifiable Medical Reasoning with Critical Evidence GraphLLM reasoning → Clinical and biomedical applications
Seedance 1.5 pro: A Native Audio-Visual Joint Generation Foundation ModelLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
Defending the Hierarchical Result Models of Precedential ConstraintLLM reasoning → Reasoning benchmarks and evaluation
neuralFOMO: Can LLMs Handle Being Second Best? Measuring Envy-Like Preferences in Multi-Agent SettingsLLM agents and tool use → Social simulation and multi-agent debate
XNNTab -- Interpretable Neural Networks for Tabular Data using Sparse AutoencodersMechanistic interpretability → Sparse autoencoder training and evaluation
From User Interface to Agent Interface: Efficiency Optimization of UI Representations for LLM AgentsLLM agents and tool use → GUI and mobile agents
RecTok: Reconstruction Distillation along Rectified FlowDiffusion models → Flow matching, time series and science
Dual-Evolving Reinforcement Learning for Automated Reward DesignDeep reinforcement learning methods → RL for games and combinatorial optimization
Automated Information Flow Selection for Multi-scenario Multi-task RecommendationMixture-of-experts → MoE for recommender systems
QoS-Aware State-Augmented Learnable Framework for 5G NR-U/Wi-Fi Coexistence: Impact of Parameter Selection and Enhanced Collision ResolutionDeep reinforcement learning methods → RL for energy, networks and operations
Dual-Phase Federated Deep Unlearning via Weight-Aware Rollback and ReconstructionLLM safety and security → Federated unlearning and poisoning
Universal Dexterous Functional Grasping via Demonstration-Editing Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
Control of a Twin Rotor using Twin Delayed Deep Deterministic Policy Gradient (TD3)Deep reinforcement learning methods → RL for robot navigation and UAV control
On the Effectiveness of Membership Inference in Targeted Data Extraction from Large Language ModelsLLM safety and security → Membership inference attacks
FROC: A Unified Framework with Risk-Optimized Control for Machine Unlearning in LLMsLLM safety and security → Machine unlearning
Security and Detectability Analysis of Unicode Text Watermarking Methods against Large Language ModelsLLM safety and security → Watermarking generated content and models
Error-Driven Prompt Optimization for Arithmetic ReasoningLLM reasoning → Mathematical reasoning
Face Identity Unlearning for Retrieval via Embedding DispersionLLM safety and security → Machine unlearning
Intrinsic-Motivation Multi-Robot Social Formation Navigation with Coordinated ExplorationDeep reinforcement learning methods → RL for robot navigation and UAV control
LINA: Learning INterventions Adaptively for Physical Alignment and Generalization in Diffusion ModelsDiffusion models → Diffusion foundations and sampling
Integrating Causal Reasoning into Automated Fact-CheckingLLM reasoning → Reasoning benchmarks and evaluation
AIR: Post-training Data Selection for Reasoning via Attention Head InfluenceLLM reasoning → Test-time compute and efficient thinking
AutoTool: Dynamic Tool Selection and Integration for Agentic ReasoningLLM reasoning → Reasoning for agents, planning and code +1
CogniEdit: Dense Gradient Flow Optimization for Fine-Grained Image EditingDiffusion models → Image generation, text-to-image and editing
Post-Training and Test-Time Scaling of Generative Agent Behavior Models for Interactive Autonomous DrivingLLM reasoning → Reinforcement learning for reasoning +1
Toward Ambulatory Vision: Learning Visually-Grounded Active View SelectionMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
STARCaster: Spatio-Temporal AutoRegressive Video Diffusion for Identity- and View-Aware Talking PortraitsDiffusion models → Audio, speech and music
Reflective Preference Optimization (RPO): Enhancing On-Policy Alignment via Hint-Guided ReflectionMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
Efficient Adaptive Rejection Sampling for Accelerating Speculative Decoding in Large Language ModelsEfficient LLM inference and serving → Speculative decoding
MMDrive: Interactive Scene Understanding Beyond Vision with Multi-representational FusionMultimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection
SACn: Soft Actor-Critic with n-step ReturnsDeep reinforcement learning methods → RL theory and sample complexity
SpeakRL: Synergizing Reasoning, Speaking, and Acting in Language Models with Reinforcement LearningLLM reasoning → Reasoning for agents, planning and code
TraPO: A Semi-Supervised Reinforcement Learning Framework for Boosting LLM ReasoningLLM reasoning → Reinforcement learning for reasoning
OXE-AugE: A Large-Scale Robot Augmentation of OXE for Scaling Cross-Embodiment Policy LearningVision-language-action models → Reinforcement learning and post-training
ADHint: Adaptive Hints with Difficulty Priors for Reinforcement LearningLLM reasoning → Reinforcement learning for reasoning +1
Sequence of Expert: Boosting Imitation Planners for Autonomous Driving through Temporal AlternationEnd-to-end autonomous driving → Imitation-learning end-to-end planners +1
Multi-Robot Motion Planning from Vision and Language using Heat-Inspired DiffusionDiffusion models → Diffusion and flow policies for robotics
Spatial-Aware VLA Pretraining through Visual-Physical Alignment from Human VideosVision-language-action models → 3D and spatial-aware VLA
Forging a Dynamic Memory: Retrieval-Guided Continual Learning for Generalist Medical Foundation ModelsMultimodal LLMs / vision-language models → Medical vision-language models +1
M-GRPO: Stabilizing Self-Supervised Reinforcement Learning for Large Language Models with Momentum-Anchored Policy OptimizationLLM reasoning → Reinforcement learning for reasoning
Deep Q-Learning-Based Intelligent Scheduling for ETL Optimization in Heterogeneous Data EnvironmentsDeep reinforcement learning methods → RL for energy, networks and operations
An Open and Reproducible Deep Research Agent for Long-Form Question AnsweringLLM agents and tool use → Web, search, and research agents
GTR-Turbo: Merged Checkpoint is Secretly a Free Teacher for Agentic VLM TrainingMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +2
Bi-Erasing: A Bidirectional Framework for Concept Removal in Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects
Motus: A Unified Latent Action World ModelWorld models → Embodied and robot world models
JoDiffusion: Jointly Diffusing Image with Pixel-Level Annotations for Semantic Segmentation PromotionDiffusion models → Medical imaging and segmentation
Few-Step Distillation for Text-to-Image Generation: A Practical GuideDiffusion models → Image generation, text-to-image and editing
Learning Terrain Aware Bipedal Locomotion via Reduced Dimensional Perceptual RepresentationsDeep reinforcement learning methods → RL for robot locomotion and manipulation
SliceMoE: Bit-Sliced Expert Caching under Miss-Rate Constraints for Efficient MoE InferenceMixture-of-experts → Efficient MoE training and inference systems
Tackling Snow-Induced Challenges: Safe Autonomous Lane-Keeping with Robust Reinforcement LearningDeep reinforcement learning methods → RL for autonomous driving and traffic
QwenLong-L1.5: Post-Training Recipe for Long-Context Reasoning and Memory ManagementLLM reasoning → Test-time compute and efficient thinking
SCAdapter: Content-Style Disentanglement for Diffusion Style TransferDiffusion models → Image generation, text-to-image and editing
Building from Scratch: A Multi-Agent Framework with Human-in-the-Loop for Multilingual Legal Terminology MappingLLM agents and tool use → Social simulation and multi-agent debate
SPAR: Session-based Pipeline for Adaptive Retrieval on Legacy File SystemsRetrieval-augmented generation → RAG pipeline design and evaluation
ProServe: Unified Multi-Priority Request Scheduling for LLM ServingEfficient LLM inference and serving → Serving systems and request scheduling

All 356