From Plato to Aristotle: Toward a General Theory of Representation Hypotheses
We show that these views can differ because the alignment criteria behind them sit at different points of one lattice, and we formalize this explanation as the Lattice Representation Hypothesis (LRH).
ProofPaper ↗
Key points
- The Platonic Representation Hypothesis (PRH) holds that increasingly capable models converge toward a common representation.
- The Aristotelian Representation Hypothesis (ARH) restricts convergence to local neighborhoods because calibration keeps the growth of neighbor overlap with model size but largely removes that of centered kernel alignment.
- We show that this removal can come from calibrating a maximum over layers, so global convergence remains open.
- We show that the spanning-tree overlap used by Structure over Geometry (SOG) mostly repeats one-nearest-neighbor overlap.
Sources (1)
- [1]From Plato to Aristotle: Toward a General Theory of Representation HypothesesarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 14, 02:41 PM
We show that these views can differ because the alignment criteria behind them sit at different points of one lattice, and we formalize this explanation as the Lattice Representation Hypothesis (LRH).
The Platonic Representation Hypothesis (PRH) holds that increasingly capable models converge toward a common representation.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 14, 2026mistralai/Timestral-209M
- Oct 14, 2026ReaDy: Recurrent States as Dynamic Context for Speculative Drafting in Hybrid Linear Attention Large Language Models
- Oct 14, 2026RuleTrek: Benchmarking Agentic Rule Reasoning in Complex Scenarios
- Oct 13, 2026Run fast, private agents on-device with LFM2.5-24B-A2B
- Oct 13, 2026Gemini 4 Argon: Built to solve your hardest problems
- Oct 13, 2026GroundGUI: GUI Grounding by High-Quality Data and Distillation-to-Reinforcement Learning