all the models — AI benchmark observatory
← Benchmarks

ProtocolQA Open-Ended

ProtocolQA Open-Ended evaluates free-form answers on troubleshooting failed experimental outcomes from common biological laboratory protocols. Kept separate from the multiple-choice ProtocolQA split.

id protocolqa-open-ended · max 1 · 1 models reported

#ModelScore

No scores for this benchmark yet.