AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

Fine-Grained GRPO for Precise Preference Alignment in Flow Models

Yujie Zhou, Pengyang Ling, Jiazi Bu, Yibin Wang, Yuhang Zang, Jiaqi Wang, Li Niu, Guangtao Zhai

CVPR 2026 · Reinforcement learning › RL from feedback · approach: Reward modeling versus direct optimization · 29 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Singular Stochastic Sampling (single-step SDE)DanceGRPO · HPSv2 test prompts (trained on HPS reward only)HPS-v2.1: 0.353 -> 0.376Improved
add Singular Stochastic SamplingDanceGRPO · HPSv2 test prompts (trained on HPS reward only)CLIP Score: 0.375 -> 0.351Hurt
add Singular Stochastic SamplingDanceGRPO · HPSv2 test prompts (trained on HPS+CLIP reward)UnifiedReward: 3.569 -> 3.688Improved
add Multi-Granularity Advantage Integration (ablation)G2RPO w/o MGAI · HPSv2 test prompts (trained on HPS+CLIP reward)UnifiedReward: 3.688 -> 3.783Improved
add Multi-Granularity Advantage Integration (ablation)G2RPO w/o MGAI · HPSv2 test prompts (trained on HPS+CLIP reward)ImageReward: 1.421 -> 1.483Improved
add Multi-Granularity Advantage Integration (ablation)G2RPO w/o MGAI · HPSv2 test prompts (trained on HPS reward only)HPS-v2.1: 0.376 -> 0.385Improved
add Multi-Granularity Advantage Integration (ablation)G2RPO w/o MGAI · HPSv2 test prompts (trained on HPS reward only)CLIP Score: 0.351 -> 0.355No clear effect
use granularities {1,2,3} instead of {1,2} (ablation)G2RPO · HPSv2 test prompts (trained on HPS+CLIP reward)HPS-v2.1: 0.375 -> 0.376No clear effect
use granularities {1,2,3} instead of single granularity {1} (ablation)G2RPO · HPSv2 test prompts (trained on HPS+CLIP reward)ImageReward: 1.421 -> 1.483Improved
use granularity {1,3} instead of {1,2,3} (ablation)G2RPO · HPSv2 test prompts (trained on HPS+CLIP reward)HPS-v2.1: 0.378 vs 0.376No clear effect
use G2RPO instead of MixGRPOMixGRPO · HPSv2 test prompts (trained on HPS+CLIP reward)UnifiedReward: 3.661 -> 3.783Improved
use G2RPO instead of MixGRPOMixGRPO · HPSv2 test prompts (trained on HPS reward only)CLIP Score: 0.358 -> 0.355Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy