Abstract
In this work, we assess the ability of foundation models to recall encyclopedic knowledge across a wide range of linguistic contexts. To support this, we: 1) produce a 20-language dataset that contains 303k factual associations paired with counterfactuals, 2) evaluate 5 models in a multilingual test, and 3) benchmark a diverse set of 24 models in an English-only test. Meta’s LLaMA achieves the highest scores in both multilingual and English-only evaluations. Yet, an analysis of LLaMA’s errors reveals significant limitations in its ability to recall facts in languages other than English, plus difficulties related to the location and gender of fact subjects. Overall, our findings suggest that today’s foundation models are far from polyglots.- Anthology ID:
- 2023.emnlp-main.691
- 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:
- 11238–11253
- Language:
- URL:
- https://aclanthology.org/2023.emnlp-main.691
- DOI:
- 10.18653/v1/2023.emnlp-main.691
- Cite (ACL):
- Tim Schott, Daniel Furman, and Shreshta Bhat. 2023. Polyglot or Not? Measuring Multilingual Encyclopedic Knowledge in Foundation Models. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 11238–11253, Singapore. Association for Computational Linguistics.
- Cite (Informal):
- Polyglot or Not? Measuring Multilingual Encyclopedic Knowledge in Foundation Models (Schott et al., EMNLP 2023)
- PDF:
- https://preview.aclanthology.org/corrections-2024-05/2023.emnlp-main.691.pdf