Robust Instant Policy: Leveraging Student's t-Regression Model for Robust In-context Imitation Learning of Robot Manipulation

TL;DR

This paper introduces a robust in-context imitation learning algorithm for robot manipulation that uses Student's t-regression to mitigate hallucination issues in large language model-based policies, significantly improving task success rates.

Contribution

The paper proposes the robust instant policy (RIP) algorithm, integrating Student's t-regression with in-context IL to enhance robustness against hallucinations in LLM-based robot policies.

Findings

RIP outperforms state-of-the-art IL methods by at least 26% in success rates.

RIP is effective in both simulated and real-world environments.

RIP particularly improves performance in low-data scenarios.

Abstract

Imitation learning (IL) aims to enable robots to perform tasks autonomously by observing a few human demonstrations. Recently, a variant of IL, called In-Context IL, utilized off-the-shelf large language models (LLMs) as instant policies that understand the context from a few given demonstrations to perform a new task, rather than explicitly updating network models with large-scale demonstrations. However, its reliability in the robotics domain is undermined by hallucination issues such as LLM-based instant policy, which occasionally generates poor trajectories that deviate from the given demonstrations. To alleviate this problem, we propose a new robust in-context imitation learning algorithm called the robust instant policy (RIP), which utilizes a Student's t-regression model to be robust against the hallucinated trajectories of instant policies to allow reliable trajectory generation. Specifically, RIP generates several candidate robot trajectories to complete a given task from an LLM and aggregates them using the Student's t-distribution, which is beneficial for ignoring outliers (i.e., hallucinations); thereby, a robust trajectory against hallucinations is generated. Our experiments, conducted in both simulated and real-world environments, show that RIP significantly outperforms state-of-the-art IL methods, with at least $26\%$ improvement in task success rates, particularly in low-data scenarios for everyday tasks. Video results available at https://sites.google.com/view/robustinstantpolicy.