文件夹: Multimodal-Generation/RL--and--Alignment

2026年5月

AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation

2026年5月

Auto-Rubric as Reward: From Implicit Preferences to Explicit Multimodal Generative Criteria

2026年5月

DiffusionOPD: A Unified Perspective of On-Policy Distillation in Diffusion Models

2026年5月

Flow-OPD: On-Policy Distillation for Flow Matching Models

2026年5月

MARBLE: Multi-Aspect Reward Balance for Diffusion RL

2026年5月

Reinforce Adjoint Matching: Scaling RL Post-Training of Diffusion and Flow-Matching Models

2026年4月

FP4 Explore, BF16 Train: Diffusion Reinforcement Learning via Efficient Rollout Scaling

2026年4月

LeapAlign - Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories

2026年4月

ParetoSlider - Diffusion Models Post-Training for Continuous Reward Control

2026年4月

V-GRPO - Online Reinforcement Learning for Denoising Generative Models Is Easier than You Think

2026年3月

Astrolabe: Steering Forward-Process Reinforcement Learning for Distilled Autoregressive Video Models

2026年3月

Calibri: Enhancing Diffusion Transformers via Parameter-Efficient Calibration

2026年3月

From Sparse to Dense: Multi-View GRPO for Flow Models via Augmented Condition Space

2026年3月

Manifold-Aware Exploration for Reinforcement Learning in Video Generation

2026年3月

UniGRPO: Unified Policy Optimization for Reasoning-Driven Visual Generation

2026年3月

Wan-R1 - Verifiable-Reinforcement Learning for Video Reasoning

2026年2月

Euphonium: Steering Video Flow Matching via Process Reward Gradient Guided Stochastic Dynamics

2026年2月

TeleBoost: A Systematic Alignment Framework for High-Fidelity, Controllable, and Robust Video Generation

2026年1月

Diffusion-DRF: Differentiable Reward Flow for Video Diffusion Fine-Tuning

2025年12月

DDRL: Data-regularized Reinforcement Learning for Diffusion Models at Scale

2025年11月

Video Generation Models Are Good Latent Reward Models

2025年9月

Advantage Weighted Matching: Aligning RL with Pretraining in Diffusion Models

2025年9月

DiffusionNFT: Online Diffusion Reinforcement with Forward Process

2025年7月

Flow Matching Policy Gradients

2025年5月

DanceGRPO: Unleashing GRPO on Visual Generation

2025年5月

Flow-GRPO: Training Flow Matching Models via Online RL

2025年2月

MJ-VIDEO: Fine-Grained Benchmarking and Rewarding Video Preferences in Video Generation

2025年1月

VideoAlign: Improving Video Generation with Human Feedback

2023年11月

Diffusion Model Alignment Using Direct Preference Optimization

2023年4月

Paper Notes

探索

文件夹: Multimodal-Generation/RL--and--Alignment

AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation

Auto-Rubric as Reward: From Implicit Preferences to Explicit Multimodal Generative Criteria

DiffusionOPD: A Unified Perspective of On-Policy Distillation in Diffusion Models

Flow-OPD: On-Policy Distillation for Flow Matching Models

MARBLE: Multi-Aspect Reward Balance for Diffusion RL

Reinforce Adjoint Matching: Scaling RL Post-Training of Diffusion and Flow-Matching Models

FP4 Explore, BF16 Train: Diffusion Reinforcement Learning via Efficient Rollout Scaling

LeapAlign - Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories

ParetoSlider - Diffusion Models Post-Training for Continuous Reward Control

V-GRPO - Online Reinforcement Learning for Denoising Generative Models Is Easier than You Think

Astrolabe: Steering Forward-Process Reinforcement Learning for Distilled Autoregressive Video Models

Calibri: Enhancing Diffusion Transformers via Parameter-Efficient Calibration

From Sparse to Dense: Multi-View GRPO for Flow Models via Augmented Condition Space

Manifold-Aware Exploration for Reinforcement Learning in Video Generation

UniGRPO: Unified Policy Optimization for Reasoning-Driven Visual Generation

Wan-R1 - Verifiable-Reinforcement Learning for Video Reasoning

Euphonium: Steering Video Flow Matching via Process Reward Gradient Guided Stochastic Dynamics

TeleBoost: A Systematic Alignment Framework for High-Fidelity, Controllable, and Robust Video Generation

Diffusion-DRF: Differentiable Reward Flow for Video Diffusion Fine-Tuning

DDRL: Data-regularized Reinforcement Learning for Diffusion Models at Scale

Video Generation Models Are Good Latent Reward Models

Advantage Weighted Matching: Aligning RL with Pretraining in Diffusion Models

DiffusionNFT: Online Diffusion Reinforcement with Forward Process

Flow Matching Policy Gradients

DanceGRPO: Unleashing GRPO on Visual Generation

Flow-GRPO: Training Flow Matching Models via Online RL

MJ-VIDEO: Fine-Grained Benchmarking and Rewarding Video Preferences in Video Generation

VideoAlign: Improving Video Generation with Human Feedback

Diffusion Model Alignment Using Direct Preference Optimization

ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation