Towards Reliable LLM-based Robot Planning via Combined Uncertainty Estimation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yin, Shiyuan, Bai, Chenjia, Zhang, Zihao, Jin, Junwei, Zhang, Xinxin, Zhang, Chi, Li, Xuelong
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916999153582080
author Yin, Shiyuan
Bai, Chenjia
Zhang, Zihao
Jin, Junwei
Zhang, Xinxin
Zhang, Chi
Li, Xuelong
author_facet Yin, Shiyuan
Bai, Chenjia
Zhang, Zihao
Jin, Junwei
Zhang, Xinxin
Zhang, Chi
Li, Xuelong
contents Large language models (LLMs) demonstrate advanced reasoning abilities, enabling robots to understand natural language instructions and generate high-level plans with appropriate grounding. However, LLM hallucinations present a significant challenge, often leading to overconfident yet potentially misaligned or unsafe plans. While researchers have explored uncertainty estimation to improve the reliability of LLM-based planning, existing studies have not sufficiently differentiated between epistemic and intrinsic uncertainty, limiting the effectiveness of uncertainty estimation. In this paper, we present Combined Uncertainty estimation for Reliable Embodied planning (CURE), which decomposes the uncertainty into epistemic and intrinsic uncertainty, each estimated separately. Furthermore, epistemic uncertainty is subdivided into task clarity and task familiarity for more accurate evaluation. The overall uncertainty assessments are obtained using random network distillation and multi-layer perceptron regression heads driven by LLM features. We validated our approach in two distinct experimental settings: kitchen manipulation and tabletop rearrangement experiments. The results show that, compared to existing methods, our approach yields uncertainty estimates that are more closely aligned with the actual execution outcomes.
format Preprint
id arxiv_https___arxiv_org_abs_2510_08044
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards Reliable LLM-based Robot Planning via Combined Uncertainty Estimation
Yin, Shiyuan
Bai, Chenjia
Zhang, Zihao
Jin, Junwei
Zhang, Xinxin
Zhang, Chi
Li, Xuelong
Robotics
Artificial Intelligence
Large language models (LLMs) demonstrate advanced reasoning abilities, enabling robots to understand natural language instructions and generate high-level plans with appropriate grounding. However, LLM hallucinations present a significant challenge, often leading to overconfident yet potentially misaligned or unsafe plans. While researchers have explored uncertainty estimation to improve the reliability of LLM-based planning, existing studies have not sufficiently differentiated between epistemic and intrinsic uncertainty, limiting the effectiveness of uncertainty estimation. In this paper, we present Combined Uncertainty estimation for Reliable Embodied planning (CURE), which decomposes the uncertainty into epistemic and intrinsic uncertainty, each estimated separately. Furthermore, epistemic uncertainty is subdivided into task clarity and task familiarity for more accurate evaluation. The overall uncertainty assessments are obtained using random network distillation and multi-layer perceptron regression heads driven by LLM features. We validated our approach in two distinct experimental settings: kitchen manipulation and tabletop rearrangement experiments. The results show that, compared to existing methods, our approach yields uncertainty estimates that are more closely aligned with the actual execution outcomes.
title Towards Reliable LLM-based Robot Planning via Combined Uncertainty Estimation
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2510.08044