Give Me the Facts! A Survey on Factual Knowledge Probing in Pre-trained Language Models

TL;DR

This survey reviews methods and datasets for probing pre-trained language models' factual knowledge, categorizes probing techniques, and discusses challenges and future directions for using PLMs as knowledge bases.

Contribution

It introduces a new categorization scheme for factual probing methods and provides a comprehensive overview of datasets and insights for future research.

Findings

Factual probing methods vary based on input/output adaptation.

Many datasets exist for evaluating factual knowledge in PLMs.

Challenges include knowledge retention and prompt optimization.

Abstract

Pre-trained Language Models (PLMs) are trained on vast unlabeled data, rich in world knowledge. This fact has sparked the interest of the community in quantifying the amount of factual knowledge present in PLMs, as this explains their performance on downstream tasks, and potentially justifies their use as knowledge bases. In this work, we survey methods and datasets that are used to probe PLMs for factual knowledge. Our contributions are: (1) We propose a categorization scheme for factual probing methods that is based on how their inputs, outputs and the probed PLMs are adapted; (2) We provide an overview of the datasets used for factual probing; (3) We synthesize insights about knowledge retention and prompt optimization in PLMs, analyze obstacles to adopting PLMs as knowledge bases and outline directions for future work.