OpenAI’s o3 Scores Less Than Half of Claimed Score in FrontierMath Test



Posted on Mon Apr 21 2025 | 12:49 pm


In December, OpenAI claimed that the o3 AI model scored 25 percent in FrontierMath, based on internal testing.




Search
Side Widget
You can put anything you want inside of these side widgets. They are easy to use, and feature the new Bootstrap 4 card containers!