Measuring how AI agents handle everyday wet-lab science
OpenAI introduces BioBasicsBench, a 100-task benchmark measuring how well AI agents interpret raw experimental data and handle everyday wet-lab science.
This is a summary curated by AIFuture. Read the complete article at the original source:
Read the full story on OpenAI