Diffusion of Thoughts: Chain-of-Thought Reasoning in Diffusion Language Models
This addresses the challenge of flexible and efficient reasoning in AI for tasks requiring complex problem-solving, representing an incremental advancement by adapting existing techniques to a new model type.
The paper tackled the problem of improving reasoning in language models by proposing Diffusion-of-Thought (DoT), which integrates diffusion models with Chain-of-Thought reasoning, resulting in a small diffusion model outperforming a much larger autoregressive model in efficiency and accuracy on tasks like multi-digit multiplication and grade school math.
Recently, diffusion models have garnered significant interest in the field of text processing due to their many potential advantages compared to conventional autoregressive models. In this work, we propose Diffusion-of-Thought (DoT), a novel approach that integrates diffusion models with Chain-of-Thought, a well-established technique for improving the reasoning ability of autoregressive language models. In contrast to autoregressive language models that make decisions in a left-to-right, token-by-token manner, DoT allows reasoning steps to diffuse over time through a diffusion language model and offers greater flexibility in trading-off computation for reasoning performance. Our experimental results demonstrate the effectiveness of DoT in multi-digit multiplication, boolean logic, and grade school math problems, with a small diffusion model outperforming a much larger autoregressive model in both efficiency and accuracy. In addition to that, 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 with diffusion language models.