一作ACL 2026Findings· Findings of ACL
D iff C o T : Diffusion-styled Chain-of-Thought Reasoning in LLM s
Shidong Cao · Hongzhan Lin · Yuxuan Gu · Ziyang Luo · Jing Ma
摘要
Chain-of-Thought (CoT) reasoning improves multi-step mathematical problem solving in large language models but remains vulnerable to exposure bias and error accumulation, as early mistakes propagate irreversibly through autoregressive decoding. In this work, we propose DiffCoT, a diffusion-styled...