AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

Reinforcing Diffusion Models by Direct Group Preference Optimization

Yihong Luo, Tianyang Hu, Jing Tang

ICLR 2026 · Reinforcement learning › RL from feedback · approach: GRPO stability and variants · 12 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
replace Flow-GRPO with DGPOSD3.5-M · GenEvaloverall score: 0.95 -> 0.97Improved
replace Flow-GRPO with DGPOSD3.5-M · visual text renderingOCR accuracy: 0.92 -> 0.96Improved
replace Flow-GRPO with DGPOSD3.5-M · human preference (PickScore)PickScore: 23.31 -> 23.89Improved
replace Flow-GRPO with DGPOSD3.5-M · DrawBench out-of-domain (GenEval-trained)Aesthetic score: 5.25 -> 5.31No clear effect
train with DGPO rather than Flow-GRPOSD3.5-M · training costtraining time: about 20x faster (near 30x on GenEval)Improved
add timestep clip strategy (ablation)DGPO · visual text renderingOCR accuracy / visual quality: 0.95 -> 0.96Improved
use ODE rollout instead of SDE rollout (ablation)DGPO · convergence and final metricsreward metricsImproved
online rollouts instead of offline (reference model) data (ablation)DGPO · reward metricsrewardImproved
use DGPO group objective instead of Diffusion DPO pairwise (ablation)DGPO · online and offline settingsrewardImproved

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