OpenAI's LifeSciBench exposes AI's science gap
The 750-task benchmark shows leading models can help with life science, but still fall far short of independent research work.
AI for Science News and Analysis
1 verified stories covering AI for Science, product updates and industry developments.