option
Home
Flash News
Content
HarperGreen
HarperGreen
July 2, 2026

OpenAI launched GeneBench-Pro, a benchmark evaluating AI models' real-world research capabilities in genomics, proteomics, and more. It tests judgment with messy, incomplete data across 129 questions in fields like statistical genetics and functional genomics. Synthetic data prevents scoring bias. OpenAI open-sourced 10 sample questions; 50 will be independently evaluated by Artificial Analysis.

OpenAI launched GeneBench-Pro, a benchmark evaluating AI models' real-world research capabilities in genomics, proteomics, and more. It tests judgment with messy, incomplete data across 129 questions in fields like statistical genetics and functional genomics. Synthetic data prevents scoring bias. OpenAI open-sourced 10 sample questions; 50 will be independently evaluated by Artificial Analysis.
Comments (0)
0/300
OR