OpenAI releases GeneBench-Pro: A research-grade benchmark for computational biology
Original title:Introducing GeneBench-Pro
AI summary
OpenAI has released GeneBench-Pro to evaluate AI agents' ability to handle ambiguity and make judgment-based analyses in computational biology. The benchmark contains 129 questions spanning 10 fields and 21 subfields, including statistical genetics and population genetics. Each question provides real, messy datasets and experimental context, requiring models to explore data, choose analysis paths, and iterate experiments. It is built with synthetic data with known complete causal structures. 82 questions have been reviewed by external domain experts to confirm their realism.
Key Takeaways
- OpenAI has released GeneBench-Pro to evaluate AI agents' ability to handle ambiguity and make…
- The benchmark contains 129 questions spanning 10 fields and 21 subfields,including statistical…
- Each question provides real,messy datasets and experimental context,requiring models to explore…
- It is built with synthetic data with known complete causal structures.
- 82 questions have been reviewed by external domain experts to confirm their realism.
Quick Facts
- Published
- Updated
- Source
- OpenAI:官网动态(RSS · 排除企业/客户案例)
- Category
- Papers & Research
- AI score
- 70
OpenAI 的新基准揭示了一个信号,GPT-5.6 在需要科学判断的模糊任务上进步神速,从不足 5% 到接近 30%,且单题成本仅几美元,这对 AI for Science 的落地想象空间影响不小。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: OpenAI releases GeneBench-Pro: A research-grade benchmark for computational biology. June 30, 2026. https://new.nat6.net/en/news/cmr0wblyn0050sl40bngva2bq/