Pre-training Language Models for Comparative Reasoning

TL;DR

This paper introduces a new pre-training framework for language models that enhances their comparative reasoning skills using scalable data collection and novel training objectives, significantly improving performance on related tasks.

Contribution

It presents a novel scalable data collection method and three new pre-training objectives specifically designed to improve comparative reasoning in language models.

Findings

Significant improvement in comparative reasoning tasks

Enhanced performance in low-resource settings

First integrated benchmark for comparative reasoning

Abstract

Comparative reasoning is a process of comparing objects, concepts, or entities to draw conclusions, which constitutes a fundamental cognitive ability. In this paper, we propose a novel framework to pre-train language models for enhancing their abilities of comparative reasoning over texts. While there have been approaches for NLP tasks that require comparative reasoning, they suffer from costly manual data labeling and limited generalizability to different tasks. Our approach introduces a novel method of collecting scalable data for text-based entity comparison, which leverages both structured and unstructured data. Moreover, we present a framework of pre-training language models via three novel objectives on comparative reasoning. Evaluation on downstream tasks including comparative question answering, question generation, and summarization shows that our pre-training framework significantly improves the comparative reasoning abilities of language models, especially under low-resource conditions. This work also releases the first integrated benchmark for comparative reasoning.