Self-CriTeach: LLM Self-Teaching and Self-Critiquing for Improving Robotic Planning via Automated Domain Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Huang, Jinbang, Li, Zhiyuan, Hu, Yuanzhao, Zhang, Zhanguang, Coates, Mark, Quan, Xingyue, Zhang, Yingxue
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916007858143232
author Huang, Jinbang
Li, Zhiyuan
Hu, Yuanzhao
Zhang, Zhanguang
Coates, Mark
Quan, Xingyue
Zhang, Yingxue
author_facet Huang, Jinbang
Li, Zhiyuan
Hu, Yuanzhao
Zhang, Zhanguang
Coates, Mark
Quan, Xingyue
Zhang, Yingxue
contents Large Language Models (LLMs) have shown strong promise for robotic task planning, particularly through the automatic generation of symbolic planning domains. However, prior work mainly treats generated domains as planning utilities. Such pipelines remain brittle under imperfect logical states and perception noise, while overlooking the potential of generated domains as scalable sources of reasoning supervision and structured reward signals. At the same time, reasoning LLMs depend on chain-of-thought (CoT) supervision, which is expensive to collect for robotic tasks, and reinforcement learning (RL) faces challenges in reward engineering. We propose Self-CriTeach, an LLM self-teaching and self-critiquing framework in which an LLM autonomously generates symbolic planning domains that serve a dual role: (1) In the self-teaching stage, generated domains are used to produce large-scale robotic planning problem--plan pairs, which are automatically converted into extended CoT trajectories for supervised fine-tuning. (2) In the self-critiquing stage, the same domains are reused as structured reward functions, providing dense feedback for reinforcement learning without manual reward engineering. This unified training pipeline yields a planning-enhanced LLM with higher planning success rates, stronger cross-task generalization, reduced inference cost, and improved resistance to imperfect logical states. GitHub Page: https://markli1hoshipu.github.io/Plan_LLM/
format Preprint
id arxiv_https___arxiv_org_abs_2509_21543
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Self-CriTeach: LLM Self-Teaching and Self-Critiquing for Improving Robotic Planning via Automated Domain Generation
Huang, Jinbang
Li, Zhiyuan
Hu, Yuanzhao
Zhang, Zhanguang
Coates, Mark
Quan, Xingyue
Zhang, Yingxue
Robotics
Large Language Models (LLMs) have shown strong promise for robotic task planning, particularly through the automatic generation of symbolic planning domains. However, prior work mainly treats generated domains as planning utilities. Such pipelines remain brittle under imperfect logical states and perception noise, while overlooking the potential of generated domains as scalable sources of reasoning supervision and structured reward signals. At the same time, reasoning LLMs depend on chain-of-thought (CoT) supervision, which is expensive to collect for robotic tasks, and reinforcement learning (RL) faces challenges in reward engineering. We propose Self-CriTeach, an LLM self-teaching and self-critiquing framework in which an LLM autonomously generates symbolic planning domains that serve a dual role: (1) In the self-teaching stage, generated domains are used to produce large-scale robotic planning problem--plan pairs, which are automatically converted into extended CoT trajectories for supervised fine-tuning. (2) In the self-critiquing stage, the same domains are reused as structured reward functions, providing dense feedback for reinforcement learning without manual reward engineering. This unified training pipeline yields a planning-enhanced LLM with higher planning success rates, stronger cross-task generalization, reduced inference cost, and improved resistance to imperfect logical states. GitHub Page: https://markli1hoshipu.github.io/Plan_LLM/
title Self-CriTeach: LLM Self-Teaching and Self-Critiquing for Improving Robotic Planning via Automated Domain Generation
topic Robotics
url https://arxiv.org/abs/2509.21543