Beneath Surface Similarity: Large Language Models Make Reasonable Scientific Analogies after Structure Abduction

TL;DR

This paper introduces a new benchmark and task for evaluating large language models' ability to perform analogical reasoning based on structural abduction in scientific contexts, revealing current limitations of models like GPT-4.

Contribution

It proposes the analogical structure abduction task and the SCAR benchmark, highlighting the gap in LLMs' understanding of relational structures in analogies.

Findings

LLMs struggle with structure abduction in scientific analogies.

Current models show limitations in reasoning about relational structures.

The SCAR benchmark provides a new standard for evaluating analogical reasoning.

Abstract

The vital role of analogical reasoning in human cognition allows us to grasp novel concepts by linking them with familiar ones through shared relational structures. Despite the attention previous research has given to word analogies, this work suggests that Large Language Models (LLMs) often overlook the structures that underpin these analogies, raising questions about the efficacy of word analogies as a measure of analogical reasoning skills akin to human cognition. In response to this, our paper introduces a task of analogical structure abduction, grounded in cognitive psychology, designed to abduce structures that form an analogy between two systems. In support of this task, we establish a benchmark called SCAR, containing 400 scientific analogies from 13 distinct fields, tailored for evaluating analogical reasoning with structure abduction. The empirical evidence underlines the continued challenges faced by LLMs, including ChatGPT and GPT-4, in mastering this task, signifying the need for future exploration to enhance their abilities.