EXPLORER: Exploration-guided Reasoning for Textual Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Basu, Kinjal, Murugesan, Keerthiram, Chaudhury, Subhajit, Campbell, Murray, Talamadupula, Kartik, Klinger, Tim
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910368984465408
author Basu, Kinjal
Murugesan, Keerthiram
Chaudhury, Subhajit
Campbell, Murray
Talamadupula, Kartik
Klinger, Tim
author_facet Basu, Kinjal
Murugesan, Keerthiram
Chaudhury, Subhajit
Campbell, Murray
Talamadupula, Kartik
Klinger, Tim
contents Text-based games (TBGs) have emerged as an important collection of NLP tasks, requiring reinforcement learning (RL) agents to combine natural language understanding with reasoning. A key challenge for agents attempting to solve such tasks is to generalize across multiple games and demonstrate good performance on both seen and unseen objects. Purely deep-RL-based approaches may perform well on seen objects; however, they fail to showcase the same performance on unseen objects. Commonsense-infused deep-RL agents may work better on unseen data; unfortunately, their policies are often not interpretable or easily transferable. To tackle these issues, in this paper, we present EXPLORER which is an exploration-guided reasoning agent for textual reinforcement learning. EXPLORER is neurosymbolic in nature, as it relies on a neural module for exploration and a symbolic module for exploitation. It can also learn generalized symbolic policies and perform well over unseen data. Our experiments show that EXPLORER outperforms the baseline agents on Text-World cooking (TW-Cooking) and Text-World Commonsense (TWC) games.
format Preprint
id arxiv_https___arxiv_org_abs_2403_10692
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle EXPLORER: Exploration-guided Reasoning for Textual Reinforcement Learning
Basu, Kinjal
Murugesan, Keerthiram
Chaudhury, Subhajit
Campbell, Murray
Talamadupula, Kartik
Klinger, Tim
Computation and Language
Artificial Intelligence
Logic in Computer Science
Text-based games (TBGs) have emerged as an important collection of NLP tasks, requiring reinforcement learning (RL) agents to combine natural language understanding with reasoning. A key challenge for agents attempting to solve such tasks is to generalize across multiple games and demonstrate good performance on both seen and unseen objects. Purely deep-RL-based approaches may perform well on seen objects; however, they fail to showcase the same performance on unseen objects. Commonsense-infused deep-RL agents may work better on unseen data; unfortunately, their policies are often not interpretable or easily transferable. To tackle these issues, in this paper, we present EXPLORER which is an exploration-guided reasoning agent for textual reinforcement learning. EXPLORER is neurosymbolic in nature, as it relies on a neural module for exploration and a symbolic module for exploitation. It can also learn generalized symbolic policies and perform well over unseen data. Our experiments show that EXPLORER outperforms the baseline agents on Text-World cooking (TW-Cooking) and Text-World Commonsense (TWC) games.
title EXPLORER: Exploration-guided Reasoning for Textual Reinforcement Learning
topic Computation and Language
Artificial Intelligence
Logic in Computer Science
url https://arxiv.org/abs/2403.10692