Online Learning of HTN Methods for integrated LLM-HTN Planning
This work addresses efficiency in AI planning systems that combine hierarchical planning with large language models, though it is incremental as it builds on the existing ChatHTN framework.
The paper tackles the problem of reducing reliance on ChatGPT for task decomposition in integrated LLM-HTN planning by learning generalized HTN methods online, resulting in fewer ChatGPT calls while maintaining or improving problem-solving success rates.
We present online learning of Hierarchical Task Network (HTN) methods in the context of integrated HTN planning and LLM-based chatbots. Methods indicate when and how to decompose tasks into subtasks. Our method learner is built on top of the ChatHTN planner. ChatHTN queries ChatGPT to generate a decomposition of a task into primitive tasks when no applicable method for the task is available. In this work, we extend ChatHTN. Namely, when ChatGPT generates a task decomposition, ChatHTN learns from it, akin to memoization. However, unlike memoization, it learns a generalized method that applies not only to the specific instance encountered, but to other instances of the same task. We conduct experiments on two domains and demonstrate that our online learning procedure reduces the number of calls to ChatGPT while solving at least as many problems, and in some cases, even more.