Algorithm of Thoughts：增强大型语言模型中的思想探索
Algorithm of Thoughts: Enhancing Exploration of Ideas in Large Language Models
AI总结：
提出 Algorithm of Thoughts 策略，利用算法推理路径和上下文示例，仅需极少查询即可高效扩展 LLMs 的思想探索，性能优于多查询树搜索方法。
摘要：
当前旨在超越“Chain-of-Thought”方法的文献，通常采用涉及停止、修改然后恢复生成过程的外部操作方式，以提升大型语言模型（LLMs）的推理能力。由于这种方式的短视性，它们增加了查询请求的数量，导致成本、内存和计算开销的增加。针对这一问题，我们提出了 Algorithm of Thoughts——一种通过算法推理路径推动 LLMs 的新策略。通过在上下文中完整使用算法示例，这种对整个过程的宏观视角利用了 LLMs 固有的递归动力学，仅需一次或几次查询即可扩展其思想探索。我们的技术优于早期的单查询方法，甚至超越了采用广泛树搜索算法的较新多查询策略，同时使用的 token 显著更少。有趣的是，我们的结果表明，使用算法指导 LLM 可以带来超越算法本身的性能，暗示了 LLM 将其直觉融入优化搜索的固有能力。我们探讨了该方法有效性的基础及其在应用中的细微差别。代码及相关内容可在以下网址找到：https://algorithm-of-thoughts.github.io。
英文摘要：
Current literature, aiming to surpass the "Chain-of-Thought" approach, often resorts to external modi operandi involving halting, modifying, and then resuming the generation process to boost Large Language Models' (LLMs) reasoning capacities. Due to their myopic perspective, they escalate the number of query requests, leading to increased costs, memory, and computational overheads. Addressing this, we propose the Algorithm of Thoughts -- a novel strategy that propels LLMs through algorithmic reasoning pathways. By employing algorithmic examples fully in-context, this overarching view of the whole process exploits the innate recurrence dynamics of LLMs, expanding their idea exploration with merely one or a few queries. Our technique outperforms earlier single-query methods and even more recent multi-query strategies that employ an extensive tree search algorithms while using significantly fewer tokens. Intriguingly, our results suggest that instructing an LLM using an algorithm can lead to performance surpassing that of the algorithm itself, hinting at LLM's inherent ability to weave its intuition into optimized searches. We probe into the underpinnings of our method's efficacy and its nuances in application. The code and related content can be found in: https://algorithm-of-thoughts.github.io.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
