[论文解读] A Framework for Neurosymbolic Robot Action Planning using Large Language Models

Alessio Capitanelli, Fulvio Mastrogiovanni|arXiv (Cornell University)|Mar 1, 2023

Machine Learning in Materials ScienceMaterials Science参考文献 28被引用 3

一句话总结

该论文提出 Teriyaki，一种神经符号框架，利用 GPT-3 逐动作生成符合 PDDL 标准的计划，实现规划与执行的并行进行。其问题求解准确率达到 95.5%，生成的计划长度最多缩短 13.5%，平均计划等待时间相比传统规划器减少 61.2%。

ABSTRACT

Symbolic task planning is a widely used approach to enforce robot autonomy due to its ease of understanding and deployment in robot architectures. However, techniques for symbolic task planning are difficult to scale in real-world, human-robot collaboration scenarios because of the poor performance in complex planning domains or when frequent re-planning is needed. We present a framework, Teriyaki, specifically aimed at bridging the gap between symbolic task planning and machine learning approaches. The rationale is training Large Language Models (LLMs), namely GPT-3, into a neurosymbolic task planner compatible with the Planning Domain Definition Language (PDDL), and then leveraging its generative capabilities to overcome a number of limitations inherent to symbolic task planners. Potential benefits include (i) a better scalability in so far as the planning domain complexity increases, since LLMs' response time linearly scales with the combined length of the input and the output, and (ii) the ability to synthesize a plan action-by-action instead of end-to-end, making each action available for execution as soon as it is generated instead of waiting for the whole plan to be available, which in turn enables concurrent planning and execution. Recently, significant efforts have been devoted by the research community to evaluate the cognitive capabilities of LLMs, with alternate successes. Instead, with Teriyaki we aim to provide an overall planning performance comparable to traditional planners in specific planning domains, while leveraging LLMs capabilities to build a look-ahead predictive planning model. Preliminary results in selected domains show that our method can: (i) solve 95.5% of problems in a test data set of 1,000 samples; (ii) produce plans up to 13.5% shorter than a traditional symbolic planner; (iii) reduce average overall waiting times for a plan availability by up to 61.4%

研究动机与目标

• 解决符号任务规划器在动态人机协作环境中面临的可扩展性与重规划低效问题。
• 通过逐动作生成计划而非端到端生成，实现规划与执行的并行。
• 利用大语言模型的生成能力，提升复杂现实领域中规划的效率与可预测性。
• 开发一种符合 PDDL 标准的神经符号规划器，在保持高准确率的同时减少规划延迟。

提出的方法

• 在传统符号规划器（Probe）生成的合成问题对上微调基于 GPT-3 的模型，以生成符合 PDDL 标准的计划。
• 采用提示工程策略，引导大语言模型输出有效的 PDDL 动作与计划结构。
• 实现模块化架构，逐个动作生成，并立即可供执行。
• 通过解耦计划生成与完整计划完成，实现规划与执行的并行。
• 利用大语言模型推理时间与输入输出长度的线性扩展特性，提升可预测性并减少延迟。
• 在 MACRO 领域上使用 1,000 个问题的测试集验证框架，以衡量准确率、计划长度与等待时间。

Figure 1 : A Baxter robot executing actions in two domains involving the manipulation of an articulated object. A human can act on an articulated object’s joint at any time, forcing the robot to re-plan.

实验结果

研究问题

• RQ1 微调后的大语言模型能否生成与传统符号规划器相当准确度的 PDDL 兼容计划？
• RQ2 与端到端规划相比，逐动作生成是否能减少计划可用的平均等待时间？
• RQ3 基于大语言模型的方法能否生成比基线符号规划器更短、更高效的计划？
• RQ4 与基于大语言模型的规划器相比，传统规划器在计划生成可预测性（以响应时间标准差衡量）方面表现如何？

主要发现

• Teriyaki 成功解决了 1,000 个测试问题中的 95.5%，准确率与最先进符号规划器相当。
• Teriyaki 生成的平均计划长度比基线符号规划器最短缩短 13.5%。
• 由于逐动作生成，计划可用的平均等待时间减少了 61.2%。
• 单个动作响应时间的标准差从 3.47 降低至 0.15，表明可预测性显著提升。
• 该框架支持规划与执行的并行进行，减少延迟，并提升人机协作中的交互流畅性。
• 尽管使用 GPT-3 的原始规划时间较慢，但该框架的生成性与可扩展性在动态、重规划密集型场景中具有显著优势。

Figure 2 : A diagram of Teriyaki fine-tuning process. Blocks in yellow represent custom code developed for data generation and processing as described in section 3.2 .