Variational Open-Domain Question Answering

Research output: Contribution to journalConference articleResearchpeer-review

  • Valentin Liévin
  • Andreas Geert Motzfeldt
  • Ida Riis Jensen
  • Winther, Ole

Retrieval-augmented models have proven to be effective in natural language processing tasks, yet there remains a lack of research on their optimization using variational inference. We introduce the Variational Open-Domain (VOD) framework for end-to-end training and evaluation of retrieval-augmented models, focusing on open-domain question answering and language modelling. The VOD objective, a self-normalized estimate of the Rényi variational bound, approximates the task marginal likelihood and is evaluated under samples drawn from an auxiliary sampling distribution (cached retriever and/or approximate posterior). It remains tractable, even for retriever distributions defined on large corpora. We demonstrate VOD's versatility by training reader-retriever BERT-sized models on multiple-choice medical exam questions. On the MedMCQA dataset, we outperform the domain-tuned Med-PaLM by +5.3% despite using 2.500× fewer parameters. Our retrieval-augmented BioLinkBERT model scored 62.9% on the MedMCQA and 55.0% on the MedQA-USMLE. Last, we show the effectiveness of our learned retriever component in the context of medical semantic search.

Original languageEnglish
JournalProceedings of Machine Learning Research
Pages (from-to)20950-20977
Number of pages28
Publication statusPublished - 2023
Event40th International Conference on Machine Learning, ICML 2023 - Honolulu, United States
Duration: 23 Jul 202329 Jul 2023


Conference40th International Conference on Machine Learning, ICML 2023
CountryUnited States

Bibliographical note

Publisher Copyright:
© 2023 Proceedings of Machine Learning Research. All rights reserved.


ID: 372185038