VERDICT verifies each reasoning step of a multimodal LLM by combining scores from three frozen judges through a closed-form consensus, reporting up to +5.95 accuracy points across six benchmarks.
Journal of the American Statistical Asso- ciation69(345), 118–121 (1974) 56
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
VERDICT: Training-Free Step-Wise Verification of Multimodal Reasoning via Disagreement-Aware Consensus
VERDICT verifies each reasoning step of a multimodal LLM by combining scores from three frozen judges through a closed-form consensus, reporting up to +5.95 accuracy points across six benchmarks.