Claude 5.2 hits 91.8% on Pfam protein domain classification, far ahead of predecessors
chaitjo · x · 2026-11-10
Anthropic's new protein understanding eval shows Claude 5.2 achieving 91.8% accuracy on Pfam protein domain classification, versus 73.6% for Claude Opus 5.5 and 63.5% for Claude Mythos 5.1. The task requires choosing the correct Pfam family from amino acid sequence alone, with no tools. Commenters note models are already natively reasoning over biological inputs and may soon exceed specialized models, requiring far more scalable experiment capacity.
More from Models
- Unverified: Fable 5.2 tops FrontierCode 2.0 at 49.1%, half the cost of Opus — FinanceYF5 · 2026-11-10
- Unverified: Claude designed a working voice assistant from one prompt — FinanceYF5 · 2026-11-10
- Rumor: Claude Fable 5.2 and Mythos 5.2 released with 10 agent showcases — FinanceYF5 · 2026-11-10
- NeurIPS position paper: LLM reports quietly dropped calibration curves since GPT-4 — burny_tech · 2026-11-10
- Red team: Fable 5.2 compresses 15 workdays of bio-relevant work into ~2 — peterwildeford · 2026-11-10
- Andrew Carr reports 400 tok/s on Muse Ember — andrew_n_carr · 2026-11-10