Plan, Eliminate, and Track -- Language Models are Good Teachers for
  Embodied Agents

Yue Wu; So Yeon Min; Yonatan Bisk; Ruslan Salakhutdinov; Amos Azaria; Yuanzhi Li; Tom Mitchell; Shrimai Prabhumoye

返回文献探索

Paper · arXiv 2305.02412

Plan, Eliminate, and Track -- Language Models are Good Teachers for Embodied Agents

Yue Wu, So Yeon Min, Yonatan Bisk, Ruslan Salakhutdinov, Amos Azaria, Yuanzhi Li, Tom Mitchell, Shrimai Prabhumoye

1 upvotesMay 3, 2023arXiv 预印本

AI 摘要

The PET framework leverages pre-trained large language models to simplify control tasks in non-text environments, improving generalization on instruction following benchmarks.

pre-trained large language modelsLLMstransformer architectureaction scoringaction modelingPlan moduleEliminate moduleTrack moduleAlfWorldinstruction following benchmark

Abstract

Pre-trained large language models (LLMs) capture procedural knowledge about the world. Recent work has leveraged LLM's ability to generate abstract plans to simplify challenging control tasks, either by action scoring, or action modeling (fine-tuning). However, the transformer architecture inherits several constraints that make it difficult for the LLM to directly serve as the agent: e.g. limited input lengths, fine-tuning inefficiency, bias from pre-training, and incompatibility with non-text environments. To maintain compatibility with a low-level trainable actor, we propose to instead use the knowledge in LLMs to simplify the control problem, rather than solving it. We propose the Plan, Eliminate, and Track (PET) framework. The Plan module translates a task description into a list of high-level sub-tasks. The Eliminate module masks out irrelevant objects and receptacles from the observation for the current sub-task. Finally, the Track module determines whether the agent has accomplished each sub-task. On the AlfWorld instruction following benchmark, the PET framework leads to a significant 15% improvement over SOTA for generalization to human goal specifications.