{
  "id": 3864,
  "url": "https://arxiv.org/abs/2605.28870v1",
  "title": "Representation Alignment Rests on Linear Structure",
  "summary": "We investigate the Platonic Representation Hypothesis (PRH) through a tripartite statistical framework of representations: signal, bias, and noise. {1) Signal:} We propose that Platonic alignment arises from the universal relationship between objects and attributes, which is encoded linearly in representations according to the Linear Representation Hypothesis (LRH). We provide evidence that LRH helps explain PRH by extracting linear object-attribute features with sparse autoencoders and showing ",
  "authors": "Kiril Bangachev, Guy Bresler, Yury Polyanskiy",
  "category": "research",
  "topics": "bias-fairness,safety-alignment,finance-investment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-05-22T12:59:01.000Z",
  "fetched_at": "2026-07-14T16:30:36.739Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/3864",
  "original_url": "https://arxiv.org/abs/2605.28870v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}