Buffer of Thoughts: Thought-Augmented Reasoning with Large Language Models
This addresses the challenge of enhancing accuracy, efficiency, and robustness in reasoning for LLM users, representing a novel method rather than an incremental improvement.
The paper tackles the problem of improving reasoning in large language models by introducing Buffer of Thoughts (BoT), a thought-augmented approach that stores and adapts high-level thought-templates from various tasks, resulting in significant performance gains such as 51% on Checkmate-in-One and 20% on Geometric Shapes, while reducing costs by 88% compared to multi-query methods.
We introduce Buffer of Thoughts (BoT), a novel and versatile thought-augmented reasoning approach for enhancing accuracy, efficiency and robustness of large language models (LLMs). Specifically, we propose meta-buffer to store a series of informative high-level thoughts, namely thought-template, distilled from the problem-solving processes across various tasks. Then for each problem, we retrieve a relevant thought-template and adaptively instantiate it with specific reasoning structures to conduct efficient reasoning. To guarantee the scalability and stability, we further propose buffer-manager to dynamically update the meta-buffer, thus enhancing the capacity of meta-buffer as more tasks are solved. We conduct extensive experiments on 10 challenging reasoning-intensive tasks, and achieve significant performance improvements over previous SOTA methods: 11% on Game of 24, 20% on Geometric Shapes and 51% on Checkmate-in-One. Further analysis demonstrate the superior generalization ability and model robustness of our BoT, while requiring only 12% of the cost of multi-query prompting methods (e.g., tree/graph of thoughts) on average. Notably, we find that our Llama3-8B+BoT has the potential to surpass Llama3-70B model. Our project is available at: https://github.com/YangLing0818/buffer-of-thought-llm