Measuring how AI agents handle everyday wet-lab science
OpenAI introduced BioBasicsBench, a 100-task benchmark for measuring how well AI agents interpret raw experimental data in wet-lab science.
Summary written automatically by AgentGid from the source's own text. Check the original before relying on it.