Measuring how AI agents handle everyday wet-lab science
Primary sourceOpenAI
OpenAI introduces BioBasicsBench, a 100-task benchmark measuring how well AI agents interpret raw experimental data and handle everyday wet-lab science.
This notice was published by OpenAI. What you see here is a brief excerpt. The full text is theirs.
Continue reading at OpenAI
