Large Language Models Are Not Strong Abstract Reasoners

TL;DR

This paper introduces a new benchmark to evaluate large language models' abstract reasoning abilities, revealing their limited performance and suggesting that guiding their generation along causal paths could enhance their reasoning skills.

Contribution

The paper presents a novel benchmark for assessing LLMs on abstract reasoning, highlighting their current limitations and proposing causal path guidance as a potential improvement.

Findings

LLMs perform poorly on abstract reasoning tasks compared to other NLP tasks.

Existing techniques do not significantly improve LLM performance on abstract reasoning.

Guiding LLMs along causal paths may enhance their reasoning and generalization abilities.

Abstract

Large Language Models have shown tremendous performance on a large variety of natural language processing tasks, ranging from text comprehension to common sense reasoning. However, the mechanisms responsible for this success remain opaque, and it is unclear whether LLMs can achieve human-like cognitive capabilities or whether these models are still fundamentally circumscribed. Abstract reasoning is a fundamental task for cognition, consisting of finding and applying a general pattern from few data. Evaluating deep neural architectures on this task could give insight into their potential limitations regarding reasoning and their broad generalisation abilities, yet this is currently an under-explored area. In this paper, we introduce a new benchmark for evaluating language models beyond memorization on abstract reasoning tasks. We perform extensive evaluations of state-of-the-art LLMs, showing that they currently achieve very limited performance in contrast with other natural language tasks, even when applying techniques that have been shown to improve performance on other NLP tasks. We argue that guiding LLM generation to follow causal paths could help improve the generalisation and reasoning abilities of LLMs.