Diffusion models have gained attention in text processing, offering many potential advantages over traditional autoregressive models. This work explores the integration of diffusion models and Chain-of-Thought (CoT), a well-established technique to improve the reasoning ability in autoregressive language models. We propose Diffusion-of-Thought (DoT), allowing reasoning steps to diffuse over time through the diffusion process. In contrast to traditional autoregressive language models that make decisions in a left-to-right, token-by-token manner, DoT offers more flexibility in the trade-off between computation and reasoning performance. Our experimental results demonstrate the effectiveness of DoT in multi-digit multiplication and grade school math problems. Additionally, DoT showcases promising self-correction abilities and benefits from existing reasoning-enhancing techniques like self-consistency decoding. Our findings contribute to the understanding and development of reasoning capabilities in diffusion language models.

翻译：扩散模型在文本处理领域备受关注，与传统自回归模型相比具有诸多潜在优势。本研究探索了扩散模型与思维链（Chain-of-Thought, CoT）这一成熟技术的融合——后者常用于提升自回归语言模型的推理能力。我们提出思维扩散（Diffusion-of-Thought, DoT）方法，允许推理步骤通过扩散过程随时间逐步展开。与传统自回归语言模型从左至右逐词决策的方式不同，DoT在计算量与推理性能的权衡中展现出更强的灵活性。实验结果表明，DoT在多位数乘法与小学数学问题中具有显著有效性。此外，DoT展现出出色的自修正能力，并能受益于自一致性解码等现有推理增强技术。本研究的发现有助于理解和开发扩散语言模型的推理能力。