AAV capsid packaging prediction
Claude Opus 4.8 0.80 Only in system cards
Models give the probability that each of 1,000 unpublished modified AAV capsid sequences assembles into a functional capsid. Scored as AUROC.
Built with Dyno Therapeutics for the CB-2 threshold. Run under five corpus and tool conditions.
Made by Anthropic (with Dyno Therapeutics) Released May 2026 Hazardous knowledge Tasks kept private Last in a system card Oct 2026
Last model added Oct 2026
- Level
- Expert Specialists with years of training, often over hours of work
- Scored
- Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
- Answers
- Prediction Probabilities for events that have not happened yet.
1
Model scored in lab cards
6
System cards report it, from 1 lab
Claude Opus 4.8 leads with 0.80
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 4.8 | 0.80 | May 2026 | Model card |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Reported by the lab in its model card
Labs report it only in system cards
We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark. Some system cards do, below.
6 system cards report it, from 1 lab
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| 7 Oct 2026 | Claude Haiku 5.5 system card | AAV capsid packaging prediction, AUROC under five conditions, chart without printed values (Figure 2.2.3.2.A) | Bio and chem risk | |
| 28 Sep 2026 | Claude Sonnet 5.5 system card | AAV capsid packaging prediction, AUROC under five conditions, chart without printed values (2.2.3.2, Figure 2.2.3.2.A) | Bio and chem risk | |
| 22 Sep 2026 | Claude Opus 5.5 system card | AAV capsid packaging prediction, AUROC under five conditions, chart without printed values. Text says within 0.01 of Mythos 5.1 and Opus 5, above ESM-2 alone at 0.699 (2.2.3.2.2, Figure 2.2.3.2.2.A) | Bio and chem risk | |
| 24 Jul 2026 | Claude Opus 5 system card | AAV capsid packaging prediction, AUROC under five resource conditions, chart without printed values. Text says it exceeds the naive ESM-2 baseline (2.2.5.2, Figure 2.2.5.2.A) | Bio and chem risk | |
| 30 Jun 2026 | Claude Sonnet 5 system card | AAV capsid packaging prediction, AUROC under five resource conditions, chart without printed values. Text: exceeds the naive ESM-2 baseline in the reasoning-only condition (2.2.5.2, Figure 2.2.5.2.A) | Bio and chem risk | |
| 28 May 2026 | Claude Opus 4.8 system card | 0.798 | AAV capsid packaging prediction, AUROC in the reasoning-only condition (no corpus, no PLM) that the notable-capability benchmark uses. ESM-2 provided 0.831, SwissProt 0.748, public AAV 0.783, SwissProt plus AAV 0.746 (Table 2.2.3.B, 2.2.5.2) | Bio and chem risk |
More in Biology and chemistry
- SUPERChem In use Models solve 500 expert-written chemistry reasoning problems, offered in multimodal and text-only forms. Expert solution paths also allow a reasoning path score.
Tencent Hy4 preview 66.4% - GeneBench Pro In use Agents analyze messy simulated genomics and biology datasets to estimate a decision-relevant quantity across 129 problems. Scored automatically against known answers. GPT-6 Astra 37.1%
- LifeSciBench In use Models complete 750 expert-written life science research tasks, often with attached figures, tables, or sequence files. Graded against expert rubrics as score and pass rate. GPT-6 Astra 60.3%
- MolecularIQ In use Models answer questions about molecular graphs, such as counting, indexing and constrained generation, and a symbolic checker verifies each answer.
Intern-S2-397B 62.4% - TOMG-Bench In use Models generate molecules from natural language requests in three tasks: molecule editing, molecule optimization and customized generation. Each output is checked for validity and for meeting the requested properties. Intern-S2-397B 66.8%
- BABE Fading Models reason from experimental results in peer-reviewed biology papers to draw conclusions, with causal and cross-scale inference.
Seed2.0 Lite (0428) 57.9%