Tests that once challenged advanced AI models are now being solved with ease, making it harder for researchers to pinpoint what current systems are actually capable of.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results