Life after BERT: What do Other Muppets Understand about Language?

Vladislav Lialin, Kevin Zhao, Namrata Shivagunde, Anna Rumshisky


Abstract
Existing pre-trained transformer analysis works usually focus only on one or two model families at a time, overlooking the variability of the architecture and pre-training objectives. In our work, we utilize the oLMpics bench- mark and psycholinguistic probing datasets for a diverse set of 29 models including T5, BART, and ALBERT. Additionally, we adapt the oLMpics zero-shot setup for autoregres- sive models and evaluate GPT networks of different sizes. Our findings show that none of these models can resolve compositional questions in a zero-shot fashion, suggesting that this skill is not learnable using existing pre-training objectives. Furthermore, we find that global model decisions such as architecture, directionality, size of the dataset, and pre-training objective are not predictive of a model’s linguistic capabilities.
Anthology ID:
2022.acl-long.227
Volume:
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
May
Year:
2022
Address:
Dublin, Ireland
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3180–3193
Language:
URL:
https://aclanthology.org/2022.acl-long.227
DOI:
10.18653/v1/2022.acl-long.227
Bibkey:
Cite (ACL):
Vladislav Lialin, Kevin Zhao, Namrata Shivagunde, and Anna Rumshisky. 2022. Life after BERT: What do Other Muppets Understand about Language?. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 3180–3193, Dublin, Ireland. Association for Computational Linguistics.
Cite (Informal):
Life after BERT: What do Other Muppets Understand about Language? (Lialin et al., ACL 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.acl-long.227.pdf
Code
 kev-zhao/life-after-bert
Data
WebText