Biobasicsbench
OpenAI Launches BioBasicsBench to Evaluate AI in Wet-Lab Science
OpenAI introduces BioBasicsBench, a 100-task benchmark to assess AI agents' ability to interpret raw experimental data in wet-lab science.