Abstract
Recent work has investigated the interesting question using pre-trained language models (PLMs) as knowledge bases for answering open questions. However, existing work is limited in using small benchmarks with high test-train overlaps. We construct a new dataset of closed-book QA using SQuAD, and investigate the performance of BART. Experiments show that it is challenging for BART to remember training facts in high precision, and also challenging to answer closed-book questions even if relevant knowledge is retained. Some promising directions are found, including decoupling the knowledge memorizing process and the QA finetune process, forcing the model to recall relevant knowledge when question answering.- Anthology ID:
- 2021.acl-long.251
- Volume:
- Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)
- Month:
- August
- Year:
- 2021
- Address:
- Online
- Venues:
- ACL | IJCNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 3241–3251
- Language:
- URL:
- https://aclanthology.org/2021.acl-long.251
- DOI:
- 10.18653/v1/2021.acl-long.251
- Cite (ACL):
- Cunxiang Wang, Pai Liu, and Yue Zhang. 2021. Can Generative Pre-trained Language Models Serve As Knowledge Bases for Closed-book QA?. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 3241–3251, Online. Association for Computational Linguistics.
- Cite (Informal):
- Can Generative Pre-trained Language Models Serve As Knowledge Bases for Closed-book QA? (Wang et al., ACL-IJCNLP 2021)
- PDF:
- https://preview.aclanthology.org/paclic-22-ingestion/2021.acl-long.251.pdf
- Code
- wangcunxiang/Can-PLM-Serve-as-KB-for-CBQA
- Data
- Natural Questions, SQuAD, TriviaQA, WebQuestions