CodeV: Code with Images for Faithful Visual Reasoning via Tool-Aware Policy Optimization
Open paper
Benchmark
Model
Rank
Results
multimodal-reasoning-on-mathvista
CodeV-7B-RL
#59
Accuracy: 71.8
multimodal-reasoning-on-mmmu
CodeV-7B-RL
#61
Accuracy: 59.3
Rank counts only results with a code link.