arXiv 2302.015602023-02cs.AI

描述、解释、规划与选择：基于大语言模型的交互式规划赋能开放世界多任务智能体

Describe, Explain, Plan and Select: Interactive Planning with Large Language Models Enables Open-World Multi-Task Agents

Institute for Artificial Intelligence, Peking University（北京大学人工智能研究院）；School of Intelligence Science and Technology, Peking University（北京大学智能科学与技术学院）；School of Computer Science, Beijing University of Posts and Telecommunications（北京邮电大学计算机学院）；Computer Science Department, University of California, Los Angeles（加利福尼亚大学洛杉矶分校计算机科学系）；Beijing Institute for General Artificial Intelligence (BIGAI)（北京通用人工智能研究院）

Zihao Wang, Shaofei Cai, Guanzhou Chen, Anji Liu, Xiaojian Ma, Yitao Liang

备注：: NeurIPS 2023

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对开放世界多任务具身智能体的长期规划与并行子目标排序难题，提出基于LLM的交互式DEPS方法，通过描述、解释和可训练目标选择器纠错并改进计划。

摘要：

我们研究了开放世界环境中多任务具身智能体的任务规划挑战。文中指出两个主要困难：1）在开放世界环境（例如 Minecraft）中执行计划，由于任务具有长期性，需要准确的多步推理；2）普通规划器在复杂计划中对并行子目标排序时，并未考虑当前智能体完成给定子任务的难易程度，因此生成的计划可能低效，甚至不可行。为此，我们提出“Describe, Explain, Plan and Select”（DEPS），这是一种基于大语言模型（LLM）的交互式规划方法。DEPS通过整合计划执行过程的 description（描述），并在扩展规划阶段遇到失败时对反馈进行 self-explanation（自我解释），从而促进对 LLM 初始生成的 plan（计划）进行更好的纠错。此外，它还包含一个目标 selector（选择器），这是一个可训练模块，根据估计的完成步骤对并行候选子目标进行排序，从而改进初始计划。我们的实验标志着首个零样本多任务智能体的里程碑：该智能体能够稳健完成 70 多个 Minecraft 任务，并使整体性能几乎翻倍。进一步测试表明，我们的方法在广泛采用的非开放式领域（即 ALFWorld 和桌面操作）中同样具有普遍有效性。消融和探索性研究详细说明了我们的设计如何优于同类方法，并展示了用该方法在 ObtainDiamond 重大挑战上取得的有前景的进展。代码发布于 https://github.com/CraftJarvis/MC-Planner。

英文摘要：

We investigate the challenge of task planning for multi-task embodied agents in open-world environments. Two main difficulties are identified: 1) executing plans in an open-world environment (e.g., Minecraft) necessitates accurate and multi-step reasoning due to the long-term nature of tasks, and 2) as vanilla planners do not consider how easy the current agent can achieve a given sub-task when ordering parallel sub-goals within a complicated plan, the resulting plan could be inefficient or even infeasible. To this end, we propose "$\underline{D}$escribe, $\underline{E}$xplain, $\underline{P}$lan and $\underline{S}$elect" ($\textbf{DEPS}$), an interactive planning approach based on Large Language Models (LLMs). DEPS facilitates better error correction on initial LLM-generated $\textit{plan}$ by integrating $\textit{description}$ of the plan execution process and providing self-$\textit{explanation}$ of feedback when encountering failures during the extended planning phases. Furthermore, it includes a goal $\textit{selector}$, which is a trainable module that ranks parallel candidate sub-goals based on the estimated steps of completion, consequently refining the initial plan. Our experiments mark the milestone of the first zero-shot multi-task agent that can robustly accomplish 70+ Minecraft tasks and nearly double the overall performances. Further testing reveals our method's general effectiveness in popularly adopted non-open-ended domains as well (i.e., ALFWorld and tabletop manipulation). The ablation and exploratory studies detail how our design beats the counterparts and provide a promising update on the $\texttt{ObtainDiamond}$ grand challenge with our approach. The code is released at https://github.com/CraftJarvis/MC-Planner.

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。

AI总结：

摘要：

英文摘要：

相关论文