用大模型实现无需任务知识的自动规划，让机器人自己拆解指令并纠错执行。
Towards Zero-Knowledge Task Planning via a Language-based Approach
- 用大模型将自然语言指令拆成子任务，生成可执行的行为树
- 执行出错时能实时调整行为树，提升任务成功率
- 无需特定任务知识，适合通用场景下的智能体行为生成
本文提出并形式化了零知识任务规划（ZKTP）问题，即在无任务特定知识的情况下制定达成目标的动作序列。我们首次探索并提出一种基于大语言模型（LLM）的方法，将自然语言指令分解为子任务，并生成可用于执行的行为树（BTs）。若执行过程中出现错误，该方法利用LLM在运行时对行为树进行动态修正，形成反馈优化循环。在AI2-THOR模拟器中的实验表明，相比依赖任务特定知识的基线方法，本方法显著提升了整体任务表现。研究验证了大模型在解决多个ZKTP关键问题上的潜力，为无需任务特定配置的自动化行为生成提供了稳健框架。
原文摘要 · Abstract (English)
In this work, we introduce and formalize the Zero-Knowledge Task Planning (ZKTP) problem, i.e., formulating a sequence of actions to achieve some goal without task-specific knowledge. Additionally, we present a first investigation and approach for ZKTP that leverages a large language model (LLM) to decompose natural language instructions into subtasks and generate behavior trees (BTs) for execution. If errors arise during task execution, the approach also uses an LLM to adjust the BTs on-the-fly in a refinement loop. Experimental validation in the AI2-THOR simulator demonstrate our approach's effectiveness in improving overall task performance compared to alternative approaches that leverage task-specific knowledge. Our work demonstrates the potential of LLMs to effectively address several aspects of the ZKTP problem, providing a robust framework for automated behavior generation with no task-specific setup.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。