Saved in:
Bibliographic Details
Main Authors: Grams, Tim, Betz, Patrick, Marton, Sascha, Lüdtke, Stefan, Bartelt, Christian
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.08925
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911118028439552
author Grams, Tim
Betz, Patrick
Marton, Sascha
Lüdtke, Stefan
Bartelt, Christian
author_facet Grams, Tim
Betz, Patrick
Marton, Sascha
Lüdtke, Stefan
Bartelt, Christian
contents Exploration is a crucial skill for in-context reinforcement learning in unknown environments. However, it remains unclear if large language models can effectively explore a partially hidden state space. This work isolates exploration as the sole objective, tasking an agent with gathering information that enhances future returns. Within this framework, we argue that measuring agent returns is not sufficient for a fair evaluation. Hence, we decompose missing rewards into their exploration and exploitation components based on the optimal achievable return. Experiments with various models reveal that most struggle to explore the state space, and weak exploration is insufficient. Nevertheless, we found a positive correlation between exploration performance and reasoning capabilities. Our decomposition can provide insights into differences in behaviors driven by prompt engineering, offering a valuable tool for refining performance in exploratory tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2501_08925
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Disentangling Exploration of Large Language Models by Optimal Exploitation
Grams, Tim
Betz, Patrick
Marton, Sascha
Lüdtke, Stefan
Bartelt, Christian
Machine Learning
Artificial Intelligence
Computation and Language
Exploration is a crucial skill for in-context reinforcement learning in unknown environments. However, it remains unclear if large language models can effectively explore a partially hidden state space. This work isolates exploration as the sole objective, tasking an agent with gathering information that enhances future returns. Within this framework, we argue that measuring agent returns is not sufficient for a fair evaluation. Hence, we decompose missing rewards into their exploration and exploitation components based on the optimal achievable return. Experiments with various models reveal that most struggle to explore the state space, and weak exploration is insufficient. Nevertheless, we found a positive correlation between exploration performance and reasoning capabilities. Our decomposition can provide insights into differences in behaviors driven by prompt engineering, offering a valuable tool for refining performance in exploratory tasks.
title Disentangling Exploration of Large Language Models by Optimal Exploitation
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2501.08925