MaestroMotif: Skill Design from Artificial Intelligence Feedback

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Klissarov, Martin, Henaff, Mikael, Raileanu, Roberta, Sodhani, Shagun, Vincent, Pascal, Zhang, Amy, Bacon, Pierre-Luc, Precup, Doina, Machado, Marlos C., D'Oro, Pierluca
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909424827760640
author Klissarov, Martin
Henaff, Mikael
Raileanu, Roberta
Sodhani, Shagun
Vincent, Pascal
Zhang, Amy
Bacon, Pierre-Luc
Precup, Doina
Machado, Marlos C.
D'Oro, Pierluca
author_facet Klissarov, Martin
Henaff, Mikael
Raileanu, Roberta
Sodhani, Shagun
Vincent, Pascal
Zhang, Amy
Bacon, Pierre-Luc
Precup, Doina
Machado, Marlos C.
D'Oro, Pierluca
contents Describing skills in natural language has the potential to provide an accessible way to inject human knowledge about decision-making into an AI system. We present MaestroMotif, a method for AI-assisted skill design, which yields high-performing and adaptable agents. MaestroMotif leverages the capabilities of Large Language Models (LLMs) to effectively create and reuse skills. It first uses an LLM's feedback to automatically design rewards corresponding to each skill, starting from their natural language description. Then, it employs an LLM's code generation abilities, together with reinforcement learning, for training the skills and combining them to implement complex behaviors specified in language. We evaluate MaestroMotif using a suite of complex tasks in the NetHack Learning Environment (NLE), demonstrating that it surpasses existing approaches in both performance and usability.
format Preprint
id arxiv_https___arxiv_org_abs_2412_08542
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle MaestroMotif: Skill Design from Artificial Intelligence Feedback
Klissarov, Martin
Henaff, Mikael
Raileanu, Roberta
Sodhani, Shagun
Vincent, Pascal
Zhang, Amy
Bacon, Pierre-Luc
Precup, Doina
Machado, Marlos C.
D'Oro, Pierluca
Artificial Intelligence
Computation and Language
Machine Learning
Describing skills in natural language has the potential to provide an accessible way to inject human knowledge about decision-making into an AI system. We present MaestroMotif, a method for AI-assisted skill design, which yields high-performing and adaptable agents. MaestroMotif leverages the capabilities of Large Language Models (LLMs) to effectively create and reuse skills. It first uses an LLM's feedback to automatically design rewards corresponding to each skill, starting from their natural language description. Then, it employs an LLM's code generation abilities, together with reinforcement learning, for training the skills and combining them to implement complex behaviors specified in language. We evaluate MaestroMotif using a suite of complex tasks in the NetHack Learning Environment (NLE), demonstrating that it surpasses existing approaches in both performance and usability.
title MaestroMotif: Skill Design from Artificial Intelligence Feedback
topic Artificial Intelligence
Computation and Language
Machine Learning
url https://arxiv.org/abs/2412.08542