OpenAI’s o3 Model Falls Short of Benchmark Claims
13
2
OpenAI’s highly anticipated reasoning model o3, along with o3-mini, was unveiled in December with claims that it could answer just over a fourth of the problems in FrontierMath, a notoriously hard set of graduate-level math puzzles. However, TechCrunch reported that the result was obtained by OpenAI on a version of o3 that used more computing power than the model released to the public. On Friday, the research institute Epoch AI, which created FrontierMath, published its own score for the public o3, revealing that it did not meet OpenAI’s benchmark claims.
By:
Previous
Why Smart Backers Are Watching These 3 Best Cryptos to Buy Under $1 for 2025
Next
Pi Network Token Unlock: 5.6 Million PI Released Today as Analyst Predicts Potential $5 Rally
