Are LLMs classical or nonmonotonic reasoners? Lessons from generics

TL;DR

This paper investigates whether large language models can perform nonmonotonic reasoning, revealing that while they mimic some human reasoning patterns, they struggle with belief stability when faced with new or conflicting information.

Contribution

The study evaluates seven state-of-the-art LLMs on nonmonotonic reasoning tasks involving generics and exceptions, highlighting their limitations in maintaining stable beliefs.

Findings

LLMs exhibit some human-like nonmonotonic reasoning patterns

They fail to maintain stable beliefs with added or conflicting information

Consistent reasoning remains challenging for LLMs

Abstract

Recent scholarship on reasoning in LLMs has supplied evidence of impressive performance and flexible adaptation to machine generated or human feedback. Nonmonotonic reasoning, crucial to human cognition for navigating the real world, remains a challenging, yet understudied task. In this work, we study nonmonotonic reasoning capabilities of seven state-of-the-art LLMs in one abstract and one commonsense reasoning task featuring generics, such as 'Birds fly', and exceptions, 'Penguins don't fly' (see Fig. 1). While LLMs exhibit reasoning patterns in accordance with human nonmonotonic reasoning abilities, they fail to maintain stable beliefs on truth conditions of generics at the addition of supporting examples ('Owls fly') or unrelated information ('Lions have manes'). Our findings highlight pitfalls in attributing human reasoning behaviours to LLMs, as well as assessing general capabilities, while consistent reasoning remains elusive.