Behavior Cloned Transformers are Neurosymbolic Reasoners

Ruoyao Wang, Peter Jansen, Marc-Alexandre Côté, Prithviraj Ammanabrolu


Abstract
In this work, we explore techniques for augmenting interactive agents with information from symbolic modules, much like humans use tools like calculators and GPS systems to assist with arithmetic and navigation. We test our agent’s abilities in text games – challenging benchmarks for evaluating the multi-step reasoning abilities of game agents in grounded, language-based environments. Our experimental study indicates that injecting the actions from these symbolic modules into the action space of a behavior cloned transformer agent increases performance on four text game benchmarks that test arithmetic, navigation, sorting, and common sense reasoning by an average of 22%, allowing an agent to reach the highest possible performance on unseen games. This action injection technique is easily extended to new agents, environments, and symbolic modules.
Anthology ID:
2023.eacl-main.204
Volume:
Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics
Month:
May
Year:
2023
Address:
Dubrovnik, Croatia
Editors:
Andreas Vlachos, Isabelle Augenstein
Venue:
EACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2777–2788
Language:
URL:
https://aclanthology.org/2023.eacl-main.204
DOI:
10.18653/v1/2023.eacl-main.204
Bibkey:
Cite (ACL):
Ruoyao Wang, Peter Jansen, Marc-Alexandre Côté, and Prithviraj Ammanabrolu. 2023. Behavior Cloned Transformers are Neurosymbolic Reasoners. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 2777–2788, Dubrovnik, Croatia. Association for Computational Linguistics.
Cite (Informal):
Behavior Cloned Transformers are Neurosymbolic Reasoners (Wang et al., EACL 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.eacl-main.204.pdf
Video:
 https://aclanthology.org/2023.eacl-main.204.mp4