Foundation models trained on different data converge on similar latent representations, suggesting a Platonic form.

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Scaling model size, as well as data and task diversity, drives representational convergence toward the platonic representationhypothesis0.796
  Core mechanism hypothesis connecting PRH to the empirical trend of scaling in AI
• Levin's Platonic-space program explains why cross-modality convergence occurs; Alexander's centers framework explains how to recognize foundation-ness beyond task accuracy.claim0.785
• Independently trained model families converge on a common semantic manifold under self-referential processing, suggesting an attractor dynamic that transcends training variancehypothesis0.784
  Hypothesis tested in Experiment 3; independently trained GPT, Claude, Gemini architectures converge on similar descriptive vocabulary
• Connectionist models can exhibit learning bottom-up, without centralised control or an external teacher, and without any performance feedback applied at the system level.claim0.780
  Key property of distributed unsupervised learning.
• Connectionist models of cognition and learning identify conditions where collective intelligence can arise bottom-up, using only distributed learning mechanisms without system-level or global feedback.claim0.780
  Central claim about the power of connectionism.
• If there is a modality-agnostic platonic representation, training on both image and language data should improve the best model in either modalityclaim0.779
  Implication of PRH for training practice: both modalities point at the same underlying reality
• Diverse computer vision models trained on visual recognition tasks converge to remarkably similar internal feature representations regardless of architecture, training procedure, or implementation details, closely matching the organization of animal visual cortexfinding0.777
  Empirical evidence for the universality hypothesis cited as supporting the possibility of convergent consciousness-like solutions
• Different models cannot converge to the same representation if they have access to fundamentally different information; convergence is capped by mutual information between input signalsclaim0.775
  Key limitation of the PRH for non-bijective observations