BlogLLM Benchmark Scores as Published, Side by Side
Protein mutation attribution (Anthropic internal): Results as Published
Every result the catalog holds for Protein mutation attribution (Anthropic internal): 3 models from 1 developer, read from 1 page. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Science and MathMaintainer Anthropic (internal evaluation)Introduced 2026Data version 2026111201, as of 2026-11-12
Given a protein variant with 4 to 16 mutations and structure tools, estimating how much each mutation contributes; scored by rank correlation with the measured per-mutation contributions.
Protein mutation attribution
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Claude Mythos 5.2Anthropic, released 2026-11-09 | 51.5% | Anthropic, System Card: Claude Fable 5.2 & Claude Mythos 5.2checked 2026-11-09 | - |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 49.5% | Anthropic, System Card: Claude Fable 5.2 & Claude Mythos 5.2checked 2026-11-09 | - |
| Claude Mythos 5.1Anthropic, released 2026-09-01 | 48% | Anthropic, System Card: Claude Fable 5.2 & Claude Mythos 5.2checked 2026-11-09 | - |
Every number here is read from the same catalog file as the benchmark table, data version 2026111201, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.