Mechanistic indicators of understanding in large language models

Bibliographic Data

ID21370950
AuthorsPierre Beckmann (0000-0001-9247-4841, Ecole polytechnique fédérale de Lausanne, corresponding author), Matthieu Queloz (0000-0001-6644-9992, University of Bern)
Year2026
Volume183
Issue6
Pages1747-1792
Publication date2026-06-01
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenuePhilosophical Studies (JOURNAL)
Journal identifiersISSN: 0031-8116 • E-ISSN: 1573-0883
PublisherSpringer Science and Business Media LLC (PUBLISHER)
DOI10.1007/s11098-026-02513-1
OpenAlexW4417275656
LanguageEN
Citations received1
References cited57

Large language models are often portrayed as merely imitating linguistic patterns without genuine understanding. We argue that recent findings in mechanistic interpretability, the emerging field probing the inner workings of LLMs, render this picture increasingly untenable—but only once those findings are integrated within a theoretical account of understanding. We propose a tiered framework for thinking about understanding in LLMs and use it to synthesize the most relevant findings to date. The framework distinguishes three hierarchical varieties of understanding, each tied to a corresponding level of computational organization: conceptual understanding emerges when a model forms “features” as directions in latent space, learning connections between diverse manifestations of a single entity or property; state-of-the-world understanding emerges when a model learns contingent factual connections between features and dynamically tracks changes in the world; principled understanding emerges when a model ceases to rely on memorized facts and discovers a compact “circuit” connecting these facts. Across these tiers, MI uncovers internal organizations that can underwrite understanding-like unification. However, these also diverge from human cognition in their parallel exploitation of heterogeneous mechanisms. Fusing philosophical theory with mechanistic evidence thus allows us to transcend binary debates over whether AI understands, paving the way for a comparative, mechanistically grounded epistemology that explores how AI understanding aligns with—and diverges from—our own

Cognition · Comprehension · Computational model · Field (mathematics) · Interpretability · Language Understanding · Computational and Text Analysis Methods · Explainable Artificial Intelligence (XAI · Language and cultural evolution

Tools

Open DOIOpen Access