AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

AlphaGRPO: Unlocking Self-Reflective Multimodal Generation in Unified Multimodal Models via Decompositional Verifiable Reward

Runhui Huang, Jie Wu, Rui Yang, Zhe Liu, Hengshuang Zhao

ICML 2026 · Image generation › Text-to-image diffusion · approach: T2I evaluation and controllable adapters

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
use DVReward instead of VIEScore holistic score (ablation)BAGEL GRPO · GenEvaloverall score: 81.7 -> 85.1Improved
use DVReward instead of VIEScore (ablation)SD3.5M FlowGRPO · TIIF-Bench Longscore: 72.9 -> 77.7Improved
use HPSv3 or UnifiedReward as reward (ablation)BAGEL GRPO · GenEvaloverall score: 84.0 -> 83.4 (HPSv3); 84.0 -> 83.7 (UnifiedReward)Hurt
use confidence score instead of binary score in DVReward (ablation)AlphaGRPO RT2I · GenEvaloverall score: 84.0 -> 85.1Improved
use confidence score instead of binary score (ablation)AlphaGRPO RT2I · TIIF-Bench Shortscore: 79.0 -> 78.9No clear effect
add False-Positive Rectification (ablation)AlphaGRPO self-reflective refinement · TIIF-Bench Longscore: 77.8 -> 79.5Improved
add perceptual quality rubrics to semantic rubrics (ablation)DVReward · TIIF-Bench Long (1024)score: 77.7 -> 79.5Improved
add perceptual quality rubrics (ablation)DVReward · GenEval (1024)overall score: 87.3 -> 87.4No clear effect
add inference-time self-reflective refinementAlphaGRPO · TIIF-Bench Overall Short (512)score: 79.1 -> 83.9Improved
train on self-reflective refinement (no editing data)AlphaGRPO · GEdit-Bench-EN overallscore: 6.52 -> 7.08Improved
train on reasoning T2I instead of self-reflective refinementAlphaGRPO · GEdit-Bench-EN overallscore: 6.52 -> 6.85Improved
train at 512 resolution, evaluate at 1024AlphaGRPO · GenEval 1024score: 86.6 -> 86.1 (AlphaGRPO); 86.6 -> 87.4 (RT2I)Mixed

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