Multimodal Graph-of-Thoughts: Hypothesis-Verification Graphs for Multimodal Reasoning in Vision-Language Models
Multimodal Graph-of-Thoughts (MM-GoT) is introduced, a training-free, verification-constrained inference framework that integrates cross-modal grounding directly into hypothesis graph search and consistently outperforms chain-of-thought reasoning baselines under matched inference budgets.
I. Belyaeva
· 0 citations