Skip to yearly menu bar Skip to main content


Poster

Algorithm of Thoughts: Enhancing Exploration of Ideas in Large Language Models

Bilgehan Sel · Ahmad Al-Tawaha · Vanshaj Khattar · Ruoxi Jia · Ming Jin

Hall C 4-9 #614
[ ] [ Project Page ] [ Paper PDF ]
Wed 24 Jul 4:30 a.m. PDT — 6 a.m. PDT

Abstract:

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

Chat is not available.