AlphaGRPO: Unlocking Self-Reflective Multimodal Generation in UMMs via Decompositional Verifiable Reward
Open paper
Benchmark
Model
Rank
Results
text-to-image-generation-on-dpg-bench
AlphaGRPO + Self-Reflective Refinement
#20
DPG-Bench Score: 87.9
image-editing-on-gedit-bench-en
AlphaGRPO
#22
Overall: 7.08
Rank counts only results with a code link.