{"total":97,"rows":[{"id":"2509.08827","title":"A Survey of Reinforcement Learning for Large Reasoning Models","created":"2025-09-10","cited_by_count":180,"area":"lr"},{"id":"2508.05629","title":"On the Generalization of SFT: A Reinforcement Learning Perspective with Reward Rectification","created":"2025-08-07","cited_by_count":127,"area":"lr"},{"id":"2508.20751","title":"Pref-GRPO: Pairwise Preference Reward-based GRPO for Stable Text-to-Image Reinforcement Learning","created":"2025-08-28","cited_by_count":92,"area":"lr"},{"id":"2509.25050","title":"Advantage Weighted Matching: Aligning RL with Pretraining in Diffusion Models","created":"2025-09-29","cited_by_count":67,"area":"diffusion"},{"id":"2509.23909","title":"EditScore: Unlocking Online RL for Image Editing via High-Fidelity Reward Modeling","created":"2025-09-28","cited_by_count":64,"area":"alignment"},{"id":"2509.06040","title":"BranchGRPO: Stable and Efficient GRPO with Structured Branching in Diffusion Models","created":"2025-09-07","cited_by_count":61,"area":"diffusion"},{"id":"2510.22319","title":"GRPO-Guard: Mitigating Implicit Over-Optimization in Flow Matching via Regulated Clipping","created":"2025-10-25","cited_by_count":49,"area":"lr"},{"id":"2601.20218","title":"DenseGRPO: From Sparse to Dense Reward for Flow Matching Model Alignment","created":"2026-01-28","cited_by_count":21,"area":"lr"},{"id":"2602.03139","title":"Diversity-Preserved Distribution Matching Distillation for Fast Visual Synthesis","created":"2026-02-03","cited_by_count":20,"area":null},{"id":"2509.23352","title":"Dynamic-TreeRPO: Breaking the Independent Trajectory Bottleneck with Structured Sampling","created":"2025-09-27","cited_by_count":18,"area":"diffusion"},{"id":"2512.24146","title":"Taming Preference Mode Collapse via Directional Decoupling Alignment in Diffusion Reinforcement Learning","created":"2025-12-30","cited_by_count":18,"area":"diffusion"},{"id":"2511.20549","title":"Flash-DMD: Towards High-Fidelity Few-Step Image Generation with Efficient Distillation and Joint Reinforcement Learning","created":"2025-11-25","cited_by_count":17,"area":"diffusion"},{"id":"2512.21514","title":"DiverseGRPO: Mitigating Mode Collapse in Image Generation via Diversity-Aware GRPO","created":"2025-12-25","cited_by_count":17,"area":"lr"},{"id":"2512.24138","title":"GARDO: Reinforcing Diffusion Models without Reward Hacking","created":"2025-12-30","cited_by_count":17,"area":"diffusion"},{"id":"2601.11087","title":"PhysRVG: Physics-Aware Unified Reinforcement Learning for Video Generative Models","created":"2026-01-16","cited_by_count":17,"area":"deeprl"},{"id":"2602.04663","title":"Rethinking the Design Space of Reinforcement Learning for Diffusion Models: On the Importance of Likelihood Estimation Beyond Loss Design","created":"2026-02-04","cited_by_count":17,"area":"lr"},{"id":"2510.25889","title":"$\\pi_\\texttt{RL}$: Online RL Fine-tuning for Flow-based Vision-Language-Action Models","created":"2025-10-29","cited_by_count":16,"area":"vla"},{"id":"2607.07675","title":"Scaling Mixture-of-Experts Video Pretraining for Embodied Intelligence","created":"2026-07-08","cited_by_count":15,"area":"moe"},{"id":"2510.21583","title":"Principled RL for Flow Matching Emerges from the Chunk-level Policy Optimization","created":"2025-10-24","cited_by_count":14,"area":"lr"},{"id":"2601.00423","title":"E-GRPO: High Entropy Steps Drive Effective Reinforcement Learning for Flow Models","created":"2026-01-01","cited_by_count":14,"area":"lr"},{"id":"2603.23500","title":"UniGRPO: Unified Policy Optimization for Reasoning-Driven Visual Generation","created":"2026-03-24","cited_by_count":13,"area":"diffusion"},{"id":"2510.02654","title":"Smart-GRPO: Smartly Sampling Noise for Efficient RL of Flow-Matching Models","created":"2025-10-03","cited_by_count":12,"area":"lr"},{"id":"2602.12529","title":"Flow-Factory: A Unified Framework for Reinforcement Learning in Flow-Matching Models","created":"2026-02-13","cited_by_count":12,"area":"diffusion"},{"id":"2511.18719","title":"Seeing What Matters: Visual Preference Policy Optimization for Visual Generation","created":"2025-11-24","cited_by_count":10,"area":"lr"},{"id":"2512.17951","title":"SuperFlow: Training Flow Matching Models with RL on the Fly","created":"2025-12-17","cited_by_count":10,"area":"diffusion"},{"id":"2601.05729","title":"TAGRPO: Boosting GRPO on Image-to-Video Generation with Direct Trajectory Alignment","created":"2026-01-09","cited_by_count":10,"area":"lr"},{"id":"2603.28718","title":"Stepwise Credit Assignment for GRPO on Flow-Matching Models","created":"2026-03-30","cited_by_count":10,"area":"diffusion"},{"id":"2508.08189","title":"Reinforcement Learning for Large Model: A Survey","created":"2025-08-11","cited_by_count":9,"area":"deeprl"},{"id":"2512.04563","title":"COOPER: A Unified Model for Cooperative Perception and Reasoning in Spatial Intelligence","created":"2025-12-04","cited_by_count":9,"area":"lr"},{"id":"2602.06422","title":"Alleviating Sparse Rewards by Modeling Step-Wise and Long-Term Sampling Effects in Flow-Based GRPO","created":"2026-02-06","cited_by_count":9,"area":"lr"},{"id":"2512.00743","title":"Multi-GRPO: Multi-Group Advantage Estimation for Text-to-Image Generation with Tree-Based Trajectories and Multiple Rewards","created":"2025-11-30","cited_by_count":8,"area":"lr"},{"id":"2604.17415","title":"Reward Score Matching: Unifying Reward-based Fine-tuning for Flow and Diffusion Models","created":"2026-04-19","cited_by_count":7,"area":"diffusion"},{"id":"2604.23380","title":"V-GRPO: Online Reinforcement Learning for Denoising Generative Models Is Easier than You Think","created":"2026-04-25","cited_by_count":7,"area":"lr"},{"id":"2605.12480","title":"OmniNFT: Modality-wise Omni Diffusion Reinforcement for Joint Audio-Video Generation","created":"2026-05-12","cited_by_count":7,"area":null},{"id":"2605.15190","title":"RAVEN: Real-time Autoregressive Video Extrapolation with Consistency-model GRPO","created":"2026-05-14","cited_by_count":7,"area":"diffusion"},{"id":"2512.12387","title":"Anchoring Values in Temporal and Group Dimensions for Flow Matching Model Alignment","created":"2025-12-13","cited_by_count":6,"area":"lr"},{"id":"2602.03425","title":"ConsistentRFT: Reducing Visual Hallucinations in Flow-based Reinforcement Fine-Tuning","created":"2026-02-03","cited_by_count":6,"area":"deeprl"},{"id":"2604.19406","title":"HP-Edit: A Human-Preference Post-Training Framework for Image Editing","created":"2026-04-21","cited_by_count":6,"area":"diffusion"},{"id":"2602.06825","title":"AEGPO: Adaptive Entropy-Guided Policy Optimization for Diffusion Models","created":"2026-02-06","cited_by_count":5,"area":"diffusion"},{"id":"2603.14128","title":"Diffusion Reinforcement Learning via Centered Reward Distillation","created":"2026-03-14","cited_by_count":5,"area":"diffusion"},{"id":"2604.25427","title":"A Systematic Post-Train Framework for Video Generation","created":"2026-04-28","cited_by_count":5,"area":"diffusion"},{"id":"2605.15803","title":"Embedding-perturbed Exploration Preference Optimization for Flow Models","created":"2026-05-15","cited_by_count":5,"area":"alignment"},{"id":"2601.15968","title":"HyperAlign: Hypernetwork for Efficient Test-Time Alignment of Diffusion Models","created":"2026-01-22","cited_by_count":4,"area":"diffusion"},{"id":"2602.01591","title":"Faster and Better Alignment for Flow Matching Models via Step-aware Advantages","created":"2026-02-02","cited_by_count":4,"area":"diffusion"},{"id":"2605.06507","title":"MARBLE: Multi-Aspect Reward Balance for Diffusion RL","created":"2026-05-07","cited_by_count":4,"area":"diffusion"},{"id":"2605.11480","title":"Efficient Adjoint Matching for Fine-tuning Diffusion Models","created":"2026-05-12","cited_by_count":4,"area":"diffusion"},{"id":"2605.15980","title":"Flash-GRPO: Efficient Alignment for Video Diffusion via One-Step Policy Optimization","created":"2026-05-15","cited_by_count":4,"area":"lr"},{"id":"2602.00413","title":"Alignment of Diffusion Model and Flow Matching for Text-to-Image Generation","created":"2026-01-31","cited_by_count":3,"area":"diffusion"},{"id":"2602.07533","title":"Joint Reward Modeling: Internalizing Chain-of-Thought for Efficient Visual Reward Models","created":"2026-02-07","cited_by_count":3,"area":"lr"},{"id":"2602.24233","title":"Enhancing Spatial Understanding in Image Generation via Reward Modeling","created":"2026-02-27","cited_by_count":3,"area":"alignment"},{"id":"2603.12648","title":"From Sparse to Dense: Multi-View GRPO for Flow Models via Augmented Condition Space","created":"2026-03-13","cited_by_count":3,"area":"lr"},{"id":"2604.06916","title":"FP4 Explore, BF16 Train: Diffusion Reinforcement Learning via Efficient Rollout Scaling","created":"2026-04-08","cited_by_count":3,"area":"diffusion"},{"id":"2604.18518","title":"UDM-GRPO: Stable and Efficient Group Relative Policy Optimization for Uniform Discrete Diffusion Models","created":"2026-04-20","cited_by_count":3,"area":"lr"},{"id":"2605.10937","title":"Power Reinforcement Post-Training of Text-to-Image Models with Super-Linear Advantage Shaping","created":"2026-05-11","cited_by_count":3,"area":"lr"},{"id":"2605.12112","title":"When Policy Entropy Constraint Fails: Preserving Diversity in Flow-based RLHF via Perceptual Entropy","created":"2026-05-12","cited_by_count":3,"area":"diffusion"},{"id":"2608.14430","title":"Designing Reinforcement Learning for Diffusion Models: A Unified Path-Space View","created":"2026-08-14","cited_by_count":3,"area":"diffusion"},{"id":"2608.23664","title":"Scaling Reinforcement Learning for Diffusion Models via Velocity Matching","created":"2026-08-24","cited_by_count":3,"area":"diffusion"},{"id":"2601.02356","title":"Talk2Move: Reinforcement Learning for Text-Instructed Object-Level Geometric Transformation in Scenes","created":"2026-01-05","cited_by_count":2,"area":"deeprl"},{"id":"2602.04928","title":"Euphonium: Steering Video Flow Matching via Process Reward Gradient Guided Stochastic Dynamics","created":"2026-02-04","cited_by_count":2,"area":"diffusion"},{"id":"2603.06623","title":"Advances in GRPO for Generation Models: A Survey","created":"2026-02-21","cited_by_count":2,"area":"lr"},{"id":"2604.06966","title":"MAR-GRPO: Stabilized GRPO for AR-diffusion Hybrid Image Generation","created":"2026-04-08","cited_by_count":2,"area":"diffusion"},{"id":"2604.14910","title":"Reward-Aware Trajectory Shaping for Few-step Visual Generation","created":"2026-04-16","cited_by_count":2,"area":"alignment"},{"id":"2604.19009","title":"Guiding Distribution Matching Distillation with Gradient-Based Reinforcement Learning","created":"2026-04-21","cited_by_count":2,"area":"diffusion"},{"id":"2604.19234","title":"Learning to Credit the Right Steps: Objective-aware Process Optimization for Visual Generation","created":"2026-04-21","cited_by_count":2,"area":"lr"},{"id":"2606.06828","title":"AdaGRPO: A Capability-Aware Adaptive Enhancement for Flow-based GRPO","created":"2026-06-05","cited_by_count":2,"area":"lr"},{"id":"2607.14962","title":"Multi-Axis Max@K Reinforcement Learning for Representative Diversity in Text-to-Image Generation","created":"2026-07-16","cited_by_count":2,"area":"deeprl"},{"id":"2512.14333","title":"Dual Attention Guided Defense Against Malicious Edits","created":"2025-12-16","cited_by_count":1,"area":"diffusion"},{"id":"2603.18528","title":"Correlation-Weighted Multi-Reward Optimization for Compositional Generation","created":"2026-03-19","cited_by_count":1,"area":"alignment"},{"id":"2603.20828","title":"EruDiff: Refactoring Knowledge in Diffusion Models for Advanced Text-to-Image Synthesis","created":"2026-03-21","cited_by_count":1,"area":"diffusion"},{"id":"2605.25661","title":"DRM: Diffusion-based Reward Model With Step-wise Guidance","created":"2026-05-25","cited_by_count":1,"area":"diffusion"},{"id":"2605.26013","title":"AdvantageFlow: Advantage-Weighted Least Squares for RL in Flow Models","created":"2026-05-25","cited_by_count":1,"area":"diffusion"},{"id":"2607.02291","title":"Optimizing Visual Generative Models via Distribution-wise Rewards","created":"2026-07-02","cited_by_count":1,"area":"alignment"},{"id":"2607.24522","title":"FlowCTS: On-policy Continuous Trajectory Supervision of Flow Models","created":"2026-07-27","cited_by_count":1,"area":"distill"},{"id":"2607.29025","title":"Evaluation-Verification Reward for Consistent Multi-Reference Image Editing","created":"2026-07-31","cited_by_count":1,"area":"alignment"},{"id":"2608.26872","title":"Self-OPD: On-Policy Distillation for Flow Matching Models without Teacher","created":"2026-08-27","cited_by_count":1,"area":"diffusion"},{"id":"2608.28010","title":"When Can Conditional Flow Matching Replace Pointwise Negative Log-Likelihood?","created":"2026-08-28","cited_by_count":1,"area":"diffusion"},{"id":"2609.04282","title":"Step Back to Move Forward: Reflection-Aware Preference Optimization for Visual Generation","created":"2026-09-03","cited_by_count":1,"area":"diffusion"},{"id":"2609.09905","title":"FlowCPO: A Unified Divergence View of Preference Alignment for Flow Models","created":"2026-09-09","cited_by_count":1,"area":"diffusion"},{"id":"2509.23339","title":"Enhancing Blind Face Restoration through Online Reinforcement Learning","created":"2025-09-27","cited_by_count":0,"area":"deeprl"},{"id":"2605.19995","title":"CogOmniControl: Reasoning-Driven Controllable Video Generation via Creative Intent Cognition","created":"2026-05-19","cited_by_count":0,"area":"lr"},{"id":"2606.00228","title":"LithoGRPO: Fast Inverse Lithography via GRPO Reinforced Flow Matching","created":"2026-05-29","cited_by_count":0,"area":"diffusion"},{"id":"2606.12072","title":"World Model Self-Distillation: Training World Models to Solve General Tasks","created":"2026-06-10","cited_by_count":0,"area":"wm"},{"id":"2606.19103","title":"ProductConsistency: Improving Product Identity Preservation in Instruction-Based Image Editing via SFT and RL","created":"2026-06-17","cited_by_count":0,"area":null},{"id":"2606.23950","title":"DivRL: Disentangled Self-Similarity Rewards for Diverse Subject-Driven Generation","created":"2026-06-22","cited_by_count":0,"area":null},{"id":"2607.07693","title":"Selective Timestep Weighting and Advantage-Based Replay for Sample-Efficient Diffusion RLHF","created":"2026-07-08","cited_by_count":0,"area":"alignment"},{"id":"2607.08056","title":"Reinforcing the Generation Order of Multimodal Masked Diffusion Models","created":"2026-07-09","cited_by_count":0,"area":"diffusion"},{"id":"2608.20011","title":"Manifold Drift in Flow Preference Optimization: A Root Cause of Reward Hacking","created":"2026-08-20","cited_by_count":0,"area":"diffusion"},{"id":"2609.03528","title":"LeanGRPO: Eliminating Redundant Recomputation in Diffusion RL","created":"2026-09-03","cited_by_count":0,"area":"deeprl"},{"id":"2609.13425","title":"ReCAST: Reward Credit Assignment across Timesteps for Online Diffusion Reinforcement","created":"2026-09-11","cited_by_count":0,"area":"diffusion"},{"id":"2609.22041","title":"$\\lambda$-Controlled GRPO: Turning Flow-Matching Ratio Instability into a Budgeted Resource","created":"2026-09-18","cited_by_count":0,"area":"lr"},{"id":"2609.32340","title":"SGA-Flow-GRPO: Spatial Gradient-Guided Credit Assignment for Flow-GRPO","created":"2026-09-26","cited_by_count":0,"area":null},{"id":"2609.38329","title":"ExploreNet: Learning Where to Explore in Diffusion GRPO","created":"2026-09-29","cited_by_count":0,"area":"lr"},{"id":"2610.00365","title":"Manifold-Constrained Initial Noise Optimization for Efficient Generative Model Alignment","created":"2026-09-30","cited_by_count":0,"area":"alignment"},{"id":"2610.02967","title":"Post-Training Frontier Text-to-Image Models by Composing Preference and Rubric Rewards","created":"2026-10-02","cited_by_count":0,"area":"alignment"},{"id":"2610.12572","title":"DiffusionART: Efficient, Stable and Scalable Reinforcement of Diffusion Models","created":"2026-10-08","cited_by_count":0,"area":"lr"},{"id":"2610.13338","title":"VideoAWR: Reinforcing Video Flow Models with Decoupled-Timestep Advantage-Weighted Rectification","created":"2026-10-09","cited_by_count":0,"area":null},{"id":"2610.20035","title":"Weighted Adjoint Matching: Finetuning Flow Matching Models with Time-Dependent Control Cost","created":"2026-10-14","cited_by_count":0,"area":"diffusion"}],"next_cursor":null}