LLMs as models for analogical reasoning

Dados Bibliográficos

ID14810643
AutoresSam Musker (0009-0000-4800-9168, Brown University, autor correspondente), Alex Duchnowski (Brown University), Raphaël Millière (0000-0001-6965-6073, Macquarie University), Ellie Pavlick (0000-0002-7155-5420, Brown University)
Ano2025
Volume145
Páginas104676
Data de publicação2025-12-01
Peer ReviewedSim
Open AccessSim
TipoARTICLE
PeriódicoJournal of Memory and Language (JOURNAL)
Identificadores do periódicoISSN: 0749-596X • E-ISSN: 1096-0821
EditoraElsevier BV (PUBLISHER)
DOI10.1016/j.jml.2025.104676
OpenAlexW4412792728
IdiomaEN
Citações recebidas1
Referências citadas19

Analogical reasoning — the capacity to identify and map structural relationships between different domains — is fundamental to human cognition and learning. Recent studies have shown that large language models (LLMs) can sometimes match humans in analogical reasoning tasks, opening the possibility that analogical reasoning might emerge from domain-general processes. However, it is still debated whether these emergent capacities are largely superficial and limited to simple relations seen during training or whether they encompass the flexible representational and mapping capabilities which are the focus of leading cognitive models of analogy. In this study, we introduce novel analogical reasoning tasks that require participants to map between semantically contentful words and sequences of letters and other abstract characters. This task necessitates the ability to flexibly re-represent rich semantic information—an ability which is known to be central to human analogy but which is thus far not well-captured by existing cognitive theories and models. We assess the performance of both human participants and LLMs on tasks focusing on reasoning from semantic structure and semantic content, introducing variations that test the robustness of their analogical inferences. Advanced LLMs match human performance across several conditions, though humans and LLMs respond differently to certain task variations and semantic distractors. Our results thus provide new evidence that LLMs might offer a how-possibly explanation of human analogical reasoning in contexts that are not yet well modeled by existing theories, but that even today’s best models are unlikely to yield how-actually explanations. • Certain LLMs match humans in some variations of a novel analogical reasoning task. • However, humans and LLMs respond differently to certain task variations. • LLMs approach human-level, but not necessarily human-like, reasoning in these tasks

Analogical reasoning · Analogy · Cognition · Cognitive psychology · Epistemology · Multi-Agent Systems and Negotiation · Natural Language Processing Techniques · Neuroscience · Psychology · Semantic Web and Ontologies