AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Alleviating Sparse Rewards by Modeling Step-Wise and Long-Term Sampling Effects in Flow-Based GRPO

Yunze Tong, Mushui Liu, Canyu Zhao, Wanggui He, Shiyi Zhang, Peng Zhang, Hongwei Zhang, Jinlong Liu, Hao Jiang

ICML 2026 · Image generation › Fast diffusion sampling · approach: Flow matching distillation · 0 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 TP-GRPO (turning-point aggregated rewards, no constraint) to Flow-GRPOFlow-GRPO on SD3.5-M · GenEvalGenEval score: 0.9673 -> 0.9714Improved
add the first-step consistency constraint to TP-GRPO (ablation)TP-GRPO on SD3.5-M · GenEvalGenEval score: 0.9714 -> 0.9725Improved
add the first-step consistency constraint to TP-GRPO (ablation)TP-GRPO on SD3.5-M · Visual Text Rendering (OCR)OCR accuracy: 0.9718 -> 0.9651Hurt
add the first-step consistency constraint to TP-GRPO (ablation)TP-GRPO on SD3.5-M · Human Preference Alignment (test prompts)PickScore: 24.73 -> 24.67No clear effect
add TP-GRPO (w constraint) to Flow-GRPOFlow-GRPO on SD3.5-M · Human Preference Alignment (test prompts)PickScore: 24.02 -> 24.67Improved
add TP-GRPO (w/o constraint) to Flow-GRPOFlow-GRPO on SD3.5-M · Visual Text Rendering (OCR)OCR accuracy: 0.9579 -> 0.9718Improved
replace TempFlow-GRPO branching process reward with TP-GRPO (w constraint)SD3.5-M on DrawBench ImageReward (text rendering policy) · DrawBench (OCR-trained policy)ImageReward: 0.6316 -> 0.7378Improved
add TP-GRPO (w/o constraint) to Flow-GRPOFlow-GRPO on SD3.5-M · DrawBench (GenEval-trained policy)ImageReward: 0.8792 -> 0.9267Improved
add TP-GRPO (w/o constraint) to Flow-GRPOFlow-GRPO on SD3.5-M · DrawBench (PickScore-trained policy)ImageReward: 1.3875 -> 1.3714No clear effect
shrink SDE sampling window moderately (to size 8) from all steps (ablation)TP-GRPO on SD3.5-M · Training curve over 2400 stepsrewardImproved
shrink SDE sampling window too far (to 4 steps) (ablation)TP-GRPO on SD3.5-M · Training curve over 2400 stepsrewardHurt
lower noise scale alpha to 0.4 from 0.7 (ablation)TP-GRPO on SD3.5-M · Training curvereward: 0.7 -> 0.4Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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