VERDICT: Training-Free Step-Wise Verification of Multimodal Reasoning via Disagreement-Aware Consensus
Open paper
Benchmark
Model
Rank
Results
vision-language-understanding-on-mmstar
VERDICT
–
Accuracy: 65.88
visual-question-answering-vqa-on-blink
VERDICT
–
Accuracy: 51.32
Rank counts only results with a code link.