ByteSized32: A Corpus and Challenge Task for Generating Task-Specific World Models Expressed as Text Games
Ruoyao Wang, Graham Todd, Xingdi Yuan, Ziang Xiao, Marc-Alexandre Côté, Peter Jansen
Abstract
In this work we investigate the capacity of language models to generate explicit, interpretable, and interactive world models of scientific and common-sense reasoning tasks. We operationalize this as a task of generating text games, expressed as hundreds of lines of Python code. To facilitate this task, we introduce ByteSized32, a corpus of 32 reasoning-focused text games totalling 20k lines of Python code. We empirically demonstrate that GPT-4 can use these games as templates for single-shot in-context learning, successfully producing runnable games on unseen topics in 28% of cases. When allowed to self-reflect on program errors, game runnability substantially increases to 58%. While evaluating simulation fidelity is labor intensive, we introduce a suite of automated metrics to assess game fidelity, technical validity, adherence to task specifications, and winnability, showing a high-degree of agreement with expert human ratings. We pose this as a challenge task to spur further development at the juncture of world modeling and code generation.- Anthology ID:
- 2023.emnlp-main.830
- Volume:
- Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
- Month:
- December
- Year:
- 2023
- Address:
- Singapore
- Editors:
- Houda Bouamor, Juan Pino, Kalika Bali
- Venue:
- EMNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 13455–13471
- Language:
- URL:
- https://aclanthology.org/2023.emnlp-main.830
- DOI:
- 10.18653/v1/2023.emnlp-main.830
- Cite (ACL):
- Ruoyao Wang, Graham Todd, Xingdi Yuan, Ziang Xiao, Marc-Alexandre Côté, and Peter Jansen. 2023. ByteSized32: A Corpus and Challenge Task for Generating Task-Specific World Models Expressed as Text Games. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 13455–13471, Singapore. Association for Computational Linguistics.
- Cite (Informal):
- ByteSized32: A Corpus and Challenge Task for Generating Task-Specific World Models Expressed as Text Games (Wang et al., EMNLP 2023)
- PDF:
- https://preview.aclanthology.org/corrections-2024-05/2023.emnlp-main.830.pdf