CoinNetwork
vip

MetaEra news, on May 13 (UTC+8), OpenAI announced the launch of a new evaluation benchmark for AI medical systems called HealthBench, co-designed by 262 doctors from 60 countries, covering 5000 real simulated dialogues. HealthBench tests the accuracy, completeness, and clinical practicality of model responses through scoring criteria set by doctors, and the code and dataset are now open.


Furthermore, OpenAI stated this morning that all Plus, Team, and Pro users can export in-depth research reports as well-formatted PDF files, including tables, images, citations, and source links. This feature is applicable to both new and old reports, and users of the Enterprise and Edu versions will have it available later. (Source: PANews)
View Original
The content is for reference only, not a solicitation or offer. No investment, tax, or legal advice provided. See Disclaimer for more risks disclosure.
  • Reward
  • Comment
  • Share
Comment
0/400
No comments