Papers

New on arXiv

Tue 30 Dec 2025

213 papers

Higress-RAG: A Holistic Optimization Framework for Enterprise Retrieval-Augmented Generation via Dual Hybrid Retrieval, Adaptive Routing, and CRAGRetrieval-augmented generation → RAG pipeline design and evaluation
GCG Attack On A Diffusion LLMDiffusion models → Robustness, privacy and adversarial aspects +1
Multi-Agent LLMs for Generating Research LimitationsLLM agents and tool use → Web, search, and research agents
GRADE: Replacing Policy Gradients with Backpropagation for LLM AlignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
LLMize: A Framework for Large Language Model-Based Numerical OptimizationLLM reasoning → Test-time compute and efficient thinking
SmartFlow Reinforcement Learning and Agentic AI for Bike-Sharing OptimisationLLM agents and tool use → Reinforcement learning for LLM agents
The Silicon Psyche: Anthropomorphic Vulnerabilities in Large Language ModelsLLM safety and security → Safety alignment, refusal and guardrails
Evaluating the Reasoning Abilities of LLMs on Underrepresented Mathematics Competition ProblemsLLM reasoning → Mathematical reasoning
Thinking on Maps: How Foundation Model Agents Explore, Remember, and Reason Map EnvironmentsLLM agents and tool use → Multimodal and embodied agents
Can Small Training Runs Reliably Guide Data Curation? Rethinking Proxy-Model PracticeLLM reasoning → Test-time compute and efficient thinking
What Drives Success in Physical Planning with Joint-Embedding Predictive World Models?World models → Latent predictive world models
HOLOGRAPH: Active Causal Discovery via Sheaf-Theoretic Alignment of Large Language Model PriorsLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
Align While Search: Belief-Guided Exploratory Inference for World-Grounded Embodied AgentsLLM agents and tool use → Multimodal and embodied agents
PackKV: Reducing KV Cache Memory Footprint through LLM-Aware Lossy CompressionEfficient LLM inference and serving → KV cache optimization
Counterfactual VLA: Self-Reflective Vision-Language-Action Model with Adaptive ReasoningLLM reasoning → Chain-of-thought reasoning for driving +2
Language Model Agents Under Attack: A Cross Model-Benchmark of Profit-Seeking Behaviors in Customer ServiceLLM agents and tool use → General LLM agent frameworks and benchmarks
DyStream: Streaming Dyadic Talking Heads Generation via Flow Matching-based Autoregressive ModelDiffusion models → Audio, speech and music
Efficient Inference for Inverse Reinforcement Learning and Dynamic Discrete Choice ModelsOffline RL and imitation learning from data → Inverse reinforcement learning +1
Lifting Vision: Ground to Aerial Localization with Reasoning Guided PlanningLLM reasoning → Spatial and embodied reasoning
Spatial-aware Vision Language Model for Autonomous DrivingMultimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection +1
SenseNova-MARS: Empowering Multimodal Agentic Reasoning and Search via Reinforcement LearningLLM reasoning → Multimodal reasoning +2
World model inspired sarcasm reasoning with large language model agentsLLM reasoning → Multimodal reasoning +1
QianfanHuijin Technical Report: A Novel Multi-Stage Training Paradigm for Finance Industrial LLMsLLM agents and tool use → Financial trading and forecasting agents
Real-world Reinforcement Learning from Suboptimal InterventionsDeep reinforcement learning methods → RL for robot locomotion and manipulation
DRL-TH: Jointly Utilizing Temporal Graph Attention and Hierarchical Fusion for UGV Navigation in Crowded EnvironmentsDeep reinforcement learning methods → RL for robot navigation and UAV control
Taming Hallucinations: Boosting MLLMs' Video Understanding via Counterfactual Video GenerationMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
RAGPart & RAGMask: Retrieval-Stage Defenses Against Corpus Poisoning in Retrieval-Augmented GenerationRetrieval-augmented generation → Privacy and security of RAG
Constrained Language Model Policy Optimization via Risk-aware Stepwise AlignmentLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
Physically-Grounded Manifold Projection Model for Generalizable Metal Artifact Reduction in Dental CBCTDiffusion models → Medical imaging and segmentation
Deep Reinforcement Learning for Solving the Fleet Size and Mix Vehicle Routing ProblemDeep reinforcement learning methods → RL for games and combinatorial optimization
MambaSeg: Harnessing Mamba for Accurate and Efficient Image-Event Semantic SegmentationState-space sequence models (S4, Mamba) → Vision Mamba backbones
Mirage: One-Step Video Diffusion for Photorealistic and Coherent Asset Editing in Driving ScenesDiffusion models → Diffusion and flow policies for robotics
GR-Dexter Technical ReportVision-language-action models → Human-video and latent-action pretraining
CorGi: Contribution-Guided Block-Wise Interval Caching for Training-Free Acceleration of Diffusion TransformersDiffusion models → Diffusion foundations and sampling
Guiding a Diffusion Transformer with the Internal Dynamics of ItselfDiffusion models → Diffusion foundations and sampling
DiffThinker: Towards Generative Multimodal Reasoning with Diffusion ModelsLLM reasoning → Multimodal reasoning +2
Deep Probabilistic Supervision for Image ClassificationKnowledge distillation → General knowledge distillation methods
Training Report of TeleChat3-MoEMixture-of-experts → Efficient MoE training and inference systems
Graph-Based Exploration for ARC-AGI-3 Interactive Reasoning TasksLLM reasoning → Multimodal reasoning
Large Emotional World ModelWorld models → World models for language agents
Taming Preference Mode Collapse via Directional Decoupling Alignment in Diffusion Reinforcement LearningDiffusion models → Diffusion and flow policies for robotics +1
Activation Steering for Masked Diffusion Language ModelsDiffusion models → Discrete and text diffusion +2
GARDO: Reinforcing Diffusion Models without Reward HackingDiffusion models → Diffusion and flow policies for robotics +2
Unified Embodied VLM Reasoning with Robotic Action via Autoregressive Discretized Pre-trainingLLM reasoning → Spatial and embodied reasoning +1
GeoBench: Rethinking Multimodal Geometric Problem-Solving via Hierarchical EvaluationLLM reasoning → Spatial and embodied reasoning
CogRec: A Cognitive Recommender Agent Fusing Large Language Models and Soar for Explainable RecommendationLLM agents and tool use → Agentic recommender systems
Guided Diffusion-based Generation of Adversarial Objects for Real-World Monocular Depth Estimation AttacksDiffusion models → Diffusion and flow policies for robotics
RainFusion2.0: Temporal-Spatial Awareness and Hardware-Efficient Block-wise Sparse AttentionDiffusion models → Video generation
LoongFlow: Directed Evolutionary Search via a Cognitive Plan-Execute-Summarize ParadigmSynthetic data and self-improvement for LLMs → Self-evolving LLM agents
How and Why LLMs Generalize: A Fine-Grained Analysis of LLM Reasoning from Cognitive Behaviors to Low-Level PatternsLLM reasoning → Reasoning benchmarks and evaluation
Policy Mirror Descent with Temporal Difference Learning: Sample Complexity under Online Markov DataDeep reinforcement learning methods → RL theory and sample complexity
AHA: Aligning Large Audio-Language Models for Reasoning Hallucinations via Counterfactual Hard NegativesLLM reasoning → Hallucination detection and mitigation +1
Jailbreaking Attacks vs. Content Safety Filters: How Far Are We in the LLM Safety Arms Race?LLM safety and security → Jailbreaking
ROAD: Reflective Optimization via Automated Debugging for Zero-Shot Agent AlignmentLLM agents and tool use → General LLM agent frameworks and benchmarks
PipeFlow: Pipelined Processing and Motion-Aware Frame Selection for Long-Form Video EditingDiffusion models → Video generation
RSAgent: Learning to Reason and Act for Text-Guided Segmentation via Multi-Turn Tool InvocationsLLM reasoning → Multimodal reasoning
FUSE-RSVLM: Feature Fusion Vision-Language Model for Remote SensingMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
On Exact Editing of Flow-Based Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
iCLP: Large Language Model Reasoning with Implicit Cognition Latent PlanningLLM reasoning → Reasoning benchmarks and evaluation
Bridging the Perception-Cognition Gap:Re-engineering SAM2 with Hilbert-Mamba for Robust VLM-based Medical DiagnosisMultimodal LLMs / vision-language models → Medical vision-language models +1
SPARK: Search Personalization via Agent-Driven Retrieval and Knowledge-sharingLLM agents and tool use → Agent memory systems
RepetitionCurse: Measuring and Understanding Router Imbalance in Mixture-of-Experts LLMs under DoS StressMixture-of-experts → Expert routing strategies
PhyAVBench: A Challenging Audio Physics-Sensitivity Benchmark for Physically Grounded Text-to-Audio-Video GenerationWorld models → Video-generation world models and simulators
Fantastic Reasoning Behaviors and Where to Find Them: Unsupervised Discovery of the Reasoning ProcessLLM reasoning → Reasoning benchmarks and evaluation +1
DriveExplorer: Images-Only Decoupled 4D Reconstruction with Progressive Restoration for Driving View ExtrapolationDiffusion models → Diffusion and flow policies for robotics
Coding With AI: From a Reflection on Industrial Practices to Future Computer Science and Software Engineering EducationLLM agents and tool use → Software engineering coding agents
Assured autonomy: How operations research powers and orchestrates generative AI systemsLLM agents and tool use → Agent threats, guardrails and governance
CEC-Zero: Zero-Supervision Character Error Correction with Self-Generated RewardsSynthetic data and self-improvement for LLMs → Self-rewarding alignment and self-training
An Comparative Analysis about KYC on a Recommendation System Toward Agentic Recommendation SystemLLM agents and tool use → Agentic recommender systems
HGMEM: Hypergraph-based Working Memory to Improve Multi-step RAG for Long-Context Complex Relational ModelingRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
Implicit geometric regularization in flow matching via density weighted Stein operatorsDiffusion models → Flow matching, time series and science
T2VAttack: Adversarial Attack on Text-to-Video Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects
Learnable Query Aggregation with KV Routing for Cross-view Geo-localisationMixture-of-experts → Vision mixture-of-experts
Soft Fitted Q-Iteration without Bellman Completeness: Occupancy Reweighting and Temperature AnnealingOffline RL and imitation learning from data → Offline RL algorithms +1
Learning to Anchor Visual Odometry: KAN-Based Pose Regression for Planetary Landing
Bielik 11B v3: Multilingual Large Language Model for European Languages
Concept Attractors in LLMs and their Applications
Closing the Data Loop: Using OpenDataArena to Engineer Superior Training Datasets
Simulation-Free PSRO: Removing Game Simulation from Policy Space Response Oracles
Automated Reproducibility Has a Problem Statement Problem
Detecting and Mitigating Treatment Leakage in Text-Based Causal Inference: Distillation and Sensitivity Analysis
ProSoftArena: Benchmarking Hierarchical Capabilities of Multimodal Agents in Professional Software Environments
LearnAD: Learning Interpretable Rules for Brain Networks in Alzheimer's Disease Classification
Quantum Machine Learning Approaches for Coordinated Stealth Attack Detection in Distributed Generation Systems
Deep versus Broad Technology Search and the Timing of Innovation Impact
Cultural Encoding in Large Language Models: The Existence Gap in AI-Mediated Brand Discovery
A-PINN: Auxiliary Physics-informed Neural Networks for Structural Vibration Analysis in Continuous Euler-Bernoulli Beam
Distribution Matching for Graph Quantification Under Structural Covariate Shift
Selective Imperfection as a Generative Framework for Analysis, Creativity and Discovery
Universal Battery Degradation Forecasting Driven by Foundation Model Across Diverse Chemistries and Conditions
Path Integral Solution for Dissipative Generative Dynamics
Harvesting AlphaEarth: Benchmarking the Geospatial Foundation Model for Agricultural Downstream Tasks
Active learning for data-driven reduced models of parametric differential systems with Bayesian operator inference
Using Large Language Models To Translate Machine Results To Human Results
Paragraph Segmentation Revisited: Towards a Standard Task for Structuring Speech
Improving the stability of the covariance-controlled adaptive Langevin thermostat for large-scale Bayesian sampling
Generalising E-prop to Deep Networks
Training-Free Color-Aware Adversarial Diffusion Sanitization for Diffusion Stegomalware Defense at Security Gateways
Automated Classification of First-Trimester Fetal Heart Views Using Ultrasound-Specific Self-Supervised Learning
F2IDiff: Real-world Image Super-resolution using Feature to Image Diffusion Foundation Model

All 213