{
  "id": 3931,
  "url": "https://arxiv.org/abs/2605.22133v3",
  "title": "Atom-level Protein Representation Learning Improves Protein Structure Prediction",
  "summary": "Recent advances in generative modeling show that pretrained representations can improve generation as conditioning features or alignment targets. Motivated by this, we study protein representations for predicting structures beyond conventional function annotation. We propose TriProRep, a structure-aware pretraining method that jointly models three aligned residue-level views: amino-acid identity, backbone geometry, and local full-atom geometry, discretely encoded via VQ-VAE tokenizers. By pretra",
  "authors": "Taewon Kim, Hyosoon Jang, Hyunjin Seo, Seonghwan Seo, Hyeongwoo Kim, Wonho Zhung et al.",
  "category": "research",
  "topics": "safety-alignment,biotech",
  "orgs": null,
  "regions": null,
  "published_at": "2026-05-21T08:07:36.000Z",
  "fetched_at": "2026-07-14T16:30:36.743Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/3931",
  "original_url": "https://arxiv.org/abs/2605.22133v3",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}