PFArena - Single-Mutant-Informed Multi-Mutant Ranking: leaderboard
Task T4 of AMix-Bio's PFArena protein-mutation benchmark (September 2026): given the measured effects of individual point mutations, rank combinations of them, which tests whether a model can reason about epistasis rather than add effects. Ground truth is deep mutational scanning data. Spearman correlation between the model's ranking and the measured combination scores (Random -0.02); higher is better. Prompted LLMs only.
Metric: Spearman. Source: huggingface.co. Saturation forecast: Around October 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6.1 Sol | 0.6 | #10 |
| 2 | GPT-6 Astra | 0.6 | #6 |
| 3 | Gemini 4 Argon | 0.58 | #11 |
| 4 | Claude Opus 5 | 0.58 | #20 |
| 5 | DeepSeek V4.1 Flash | 0.57 | #58 |
| 6 | Gemini 3.1 Pro (Preview) | 0.53 | #91 |
| 7 | DeepSeek V4 Pro | 0.53 | #181 |
| 8 | GLM-5.2 | 0.5 | #137 |
| 9 | Kimi K3 | 0.48 | #53 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=pfarena-single-mutant-informed-multi-mutant-ranking · How It Works · Data refreshed daily, snapshot 2026-11-12.