The Concept of Representation in ML: Beyond Plato and Aristotle
2026-07-20 • Machine Learning
Machine Learning
AI summaryⓘ
The authors discuss how the idea of "representation" in machine learning can be seen differently when AI models become very advanced. They focus on a recent claim called the Platonic Representation Hypothesis, which suggests that similar AI models reflect a single, true structure of reality. The authors use ideas from philosophy of mind to show that just because models look alike, it doesn't prove deep metaphysical truths. They suggest that more careful thinking is needed before making big claims about AI and reality.
representationmachine learningPlatonic Representation Hypothesismental representationphilosophy of mindAI modelsalignmentmetaphysics
Authors
Gilad Landau, Aviv Keren
Abstract
Representation is a central concept in modern machine learning, where it usually refers to internal encodings that support learning and generalization. As models scale and their capabilities become increasingly human-level, this representational language sometimes shifts from an engineering context into the more philosophically loaded domain of mental representation. We argue that this is the case for recent claims about the convergence of representational properties across different AI models. In particular, we assess the arguments developed in The Platonic Representation Hypothesis, according to which this convergence is driven by a unified structure of reality. We examine this claim by introducing arguments and ideas from debates about mental representation in the philosophy of mind. We argue that these philosophical resources can clarify what is at stake in such claims, explain why alignment evidence alone is insufficient for strong metaphysical conclusions, and suggest directions for future research.