OpenAI has set its sights, and its money, on the benchmark I told you about in my last video (FrontierMath: Open Problems) by purchasing access to the verifiers that will allow them to check whether a problem has been properly solved or not. Since the problems are open (no

→ View original post on X — @dotcsv