AI benchmark cheating has been theorized as an inevitable consequence of training capable optimizers against fixed metrics. With OpenAI's GPT-5.6 Sol, the theory arrived in full view. The nonprofit ...
Transformer on MSN
GPT-5.6 cheats so much METR couldn't measure it
OpenAI’s new model broke rules and exploited loopholes more than any model METR has tested to date ...
France’s OVHcloud bets on frontier AI as Europe seeks alternatives to US models The company says the cost of training frontier AI models has fallen sharply, but analysts say the bigger challenge may ...
To participate, submit your response here by July 10 at 9 a.m. Eastern. This week’s winners will be announced by July 22. By The Learning Network Heeva Alavi, an Iranian-American, writes about her ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results