Probing Linguistic Information For Logical Inference In Pre-trained Language Models

TL;DR

This paper investigates the extent to which pre-trained language models encode linguistic information necessary for logical inference, revealing their strengths, weaknesses, and potential to support symbolic inference methods.

Contribution

It introduces a probing methodology and datasets to evaluate linguistic knowledge in language models for logical inference, highlighting their capabilities and limitations.

Findings

Language models encode several types of linguistic information for inference.

Some linguistic information is weakly encoded in pre-trained models.

Fine-tuning improves the encoding of missing linguistic information.

Abstract

Progress in pre-trained language models has led to a surge of impressive results on downstream tasks for natural language understanding. Recent work on probing pre-trained language models uncovered a wide range of linguistic properties encoded in their contextualized representations. However, it is unclear whether they encode semantic knowledge that is crucial to symbolic inference methods. We propose a methodology for probing linguistic information for logical inference in pre-trained language model representations. Our probing datasets cover a list of linguistic phenomena required by major symbolic inference systems. We find that (i) pre-trained language models do encode several types of linguistic information for inference, but there are also some types of information that are weakly encoded, (ii) language models can effectively learn missing linguistic information through fine-tuning. Overall, our findings provide insights into which aspects of linguistic information for logical inference do language models and their pre-training procedures capture. Moreover, we have demonstrated language models' potential as semantic and background knowledge bases for supporting symbolic inference methods.