LLMs Can Plan Only If We Tell Them

TL;DR

This paper demonstrates that large language models can generate effective long-horizon plans autonomously when enhanced with specific algorithmic improvements, surpassing previous methods and human performance on planning benchmarks.

Contribution

The authors introduce AoT+ enhancements to Algorithm-of-Thoughts, enabling LLMs to independently produce competitive long-term plans without external feedback.

Findings

Achieved state-of-the-art results in planning benchmarks

Outperformed prior methods and human baselines

Enabled autonomous long-horizon planning with LLMs

Abstract

Large language models (LLMs) have demonstrated significant capabilities in natural language processing and reasoning, yet their effectiveness in autonomous planning has been under debate. While existing studies have utilized LLMs with external feedback mechanisms or in controlled environments for planning, these approaches often involve substantial computational and development resources due to the requirement for careful design and iterative backprompting. Moreover, even the most advanced LLMs like GPT-4 struggle to match human performance on standard planning benchmarks, such as the Blocksworld, without additional support. This paper investigates whether LLMs can independently generate long-horizon plans that rival human baselines. Our novel enhancements to Algorithm-of-Thoughts (AoT), which we dub AoT+, help achieve state-of-the-art results in planning benchmarks out-competing prior methods and human baselines all autonomously.