M$^3$CoT

M$^3$CoT is a novel benchmark that contains multi-domain, multi-step, and multi-modal tasks designed to evaluate the reasoning capabilities of models in the context of Multi-modal Chain-of-Thought (MCoT).

Papers using M$^3$CoT (3)

Reasoning in the Dark: Interleaved Vision-Text Reasoning in Latent Space2025 · 7 citesEvolutionary Prompt Optimization Discovers Emergent Multimodal Reasoning Strategies in Vision-Language Models2025M$^3$CoT: A Novel Benchmark for Multi-Domain Multi-step Multi-modal Chain-of-Thought2024