arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.11489cs.CL
收藏

计划是什么?评估与开发面向语言模型的规划感知技术

What's the Plan? Evaluating and Developing Planning-Aware Techniques for Language Models

IBM Research(IBM研究院);Bar-Ilan University(巴伊兰大学)

Eran Hirsch, Guy Uziel, Ateret Anaby-Tavor

AI总结:

本文通过实验揭示大型语言模型在规划任务上的不足,提出结合经典规划方法的混合方法SimPlan,并在多个规划领域验证其显著优于现有LLM规划器。

摘要:

规划是人工智能中的一项基本任务,涉及在给定环境中找到一系列实现特定目标的动作。大型语言模型(LLMs)越来越多地被用于需要规划能力的应用,例如网络或具身智能体。与近期研究一致,我们通过实验证明,大型语言模型缺乏规划所需的必要技能。基于这些观察,我们主张采用一种将大型语言模型与经典规划方法相结合的混合方法的潜力。然后,我们介绍了SimPlan,一种新颖的混合方法,并在一个新的具有挑战性的设置中评估其性能。我们在各种规划领域的大量实验表明,SimPlan显著优于现有的基于大型语言模型的规划器。

英文摘要:

Planning is a fundamental task in artificial intelligence that involves finding a sequence of actions that achieve a specified goal in a given environment. Large language models (LLMs) are increasingly used for applications that require planning capabilities, such as web or embodied agents. In line with recent studies, we demonstrate through experimentation that LLMs lack necessary skills required for planning. Based on these observations, we advocate for the potential of a hybrid approach that combines LLMs with classical planning methodology. Then, we introduce SimPlan, a novel hybrid-method, and evaluate its performance in a new challenging setup. Our extensive experiments across various planning domains demonstrate that SimPlan significantly outperforms existing LLM-based planners.

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。

↑