一作ACL 2026Poster· Natural Language Processing
M ath C anvas: Intrinsic Visual Chain-of-Thought for Multimodal Mathematical Reasoning
Weikang Shi · Aldrich Yu · Rongyao Fang · Houxing Ren · Ke Wang · Aojun Zhou · Changyao Tian · Xinyu Fu · Yuxuan Hu · Zimu Lu · Linjiang Huang · Si Liu · Rui Liu · Hongsheng Li
摘要
While Large Language Models (LLMs) have excelled in textual reasoning, they struggle with mathematical domains like geometry that intrinsically rely on visual aids. Existing approaches to Visual Chain-of-Thought (VCoT) are often limited by rigid external tools or fail to generate the high-fidelit...