Reconstruction Alignment Improves Unified Multimodal Models

Open paper
Benchmark Model Rank Results
text-to-image-generation-on-dpg-benchHarmon-1.5B-RecA-plus (extra data)#10DPG-Bench Score: 88.15
text-to-image-generation-on-genevalHarmon-1.5B-RecA-plus (extra data)#20Overall: 0.90

Rank counts only results with a code link.