lilGym: Natural Language Visual Reasoning with Reinforcement Learning

Anne Wu, Kiante Brantley, Noriyuki Kojima, Yoav Artzi


Abstract
We present lilGym, a new benchmark for language-conditioned reinforcement learning in visual environments. lilGym is based on 2,661 highly-compositional human-written natural language statements grounded in an interactive visual environment. We introduce a new approach for exact reward computation in every possible world state by annotating all statements with executable Python programs. Each statement is paired with multiple start states and reward functions to form thousands of distinct Markov Decision Processes of varying difficulty. We experiment with lilGym with different models and learning regimes. Our results and analysis show that while existing methods are able to achieve non-trivial performance, lilGym forms a challenging open problem. lilGym is available at https://lil.nlp.cornell.edu/lilgym/.
Anthology ID:
2023.acl-long.512
Volume:
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
July
Year:
2023
Address:
Toronto, Canada
Editors:
Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
9214–9234
Language:
URL:
https://aclanthology.org/2023.acl-long.512
DOI:
10.18653/v1/2023.acl-long.512
Bibkey:
Cite (ACL):
Anne Wu, Kiante Brantley, Noriyuki Kojima, and Yoav Artzi. 2023. lilGym: Natural Language Visual Reasoning with Reinforcement Learning. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 9214–9234, Toronto, Canada. Association for Computational Linguistics.
Cite (Informal):
lilGym: Natural Language Visual Reasoning with Reinforcement Learning (Wu et al., ACL 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/ingest-acl-2023-videos/2023.acl-long.512.pdf