Abstract
We analyze the masked language modeling pretraining objective function from the perspective of the Distributional Hypothesis. We investigate whether the better sample efficiency and the better generalization capability of models pretrained with masked language modeling can be attributed to the semantic similarity encoded in the pretraining data’s distributional property. Via a synthetic dataset, our analysis suggests that distributional property indeed leads to the better sample efficiency of pretrained masked language models, but does not fully explain the generalization capability. We also conduct an analysis over two real-world datasets and demonstrate that the distributional property does not explain the generalization ability of pretrained natural language models either. Our results illustrate our limited understanding of model pretraining and provide future research directions.- Anthology ID:
- 2023.emnlp-main.637
- 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:
- 10305–10321
- Language:
- URL:
- https://aclanthology.org/2023.emnlp-main.637
- DOI:
- 10.18653/v1/2023.emnlp-main.637
- Cite (ACL):
- Ting-Rui Chiang and Dani Yogatama. 2023. The Distributional Hypothesis Does Not Fully Explain the Benefits of Masked Language Model Pretraining. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 10305–10321, Singapore. Association for Computational Linguistics.
- Cite (Informal):
- The Distributional Hypothesis Does Not Fully Explain the Benefits of Masked Language Model Pretraining (Chiang & Yogatama, EMNLP 2023)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-1/2023.emnlp-main.637.pdf