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 CoT framework that reformulates CoT reasoning as an iterative denoising process. DiffCoT integrates diffusion principles at the reasoning-step level via a sliding-window mechanism, enabling unified generation and retrospective correction of intermediate steps while preserving token-level autoregression. To maintain causal consistency, we further introduce a causal diffusion noise schedule that respects the temporal structure of reasoning chains. Extensive experiments on three multi-step CoT reasoning benchmarks across diverse model backbones demonstrate that DiffCoT consistently outperforms existing CoT preference optimization methods, yielding improved robustness and error-correction capability in CoT reasoning.
@article{arxiv.2601.03559,
title = {DiffCoT: Diffusion-styled Chain-of-Thought Reasoning in LLMs},
author = {Shidong Cao and Hongzhan Lin and Yuxuan Gu and Ziyang Luo and Jing Ma},
journal= {arXiv preprint arXiv:2601.03559},
year = {2026}
}
Comments
DiffCoT improves multi-step LLM reasoning by applying diffusion-based iterative denoising to correct intermediate Chain-of-Thought steps