OpenAI News·· 21 h ago
Measuring how AI agents handle everyday wet-lab science
Measuring how AI agents handle everyday wet-lab science
AI summary
OpenAI introduces BioBasicsBench, a 100-task benchmark measuring how well AI agents interpret raw experimental data and handle everyday wet-lab science.
Selection record
Threshold 60Official, first-handFirst 38Second 38
Not admittedSum of both 76 < twice the threshold 120
- Source tier
- Official, first-hand; this tier's threshold is 60
- Pre-filter
- passed：OpenAI发布AI智能体湿实验室基准
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: OpenAI News · openai.com