How Context Affects Language Models' Factual PredictionsDownload PDF

14 Feb 2020 (modified: 10 May 2020)AKBC 2020 Conference Blind SubmissionReaders: Everyone
  • Subject Areas: QuestionAnswering and Reasoning
  • Archival Status: Archival
  • Abstract: When pre-trained on large unsupervised textual corpora, language models are able to store and retrieve factual knowledge to some extent, making it possible to use them directly for zero-shot cloze-style question answering. However, storing factual knowledge in a fixed number of weights of a language model clearly has limitations. Previous approaches have successfully provided access to information outside the model weights using supervised architectures that combine an information retrieval system with a machine reading component. In this paper, we go one step further and integrate information from a retrieval system with a pre-trained language model in a purely unsupervised way. We report that augmenting pre-trained language models in this way dramatically improves performance and that it is competitive with a supervised machine reading baseline without requiring any supervised training. Furthermore, processing query and context with different segment tokens allows BERT to utilize its Next Sentence Prediction pre-trained classifier to determine whether the context is relevant or not, substantially improving BERT's zero-shot cloze-style question-answering performance and making its predictions robust to noisy contexts.
7 Replies

Loading