3 ms·

So as of a month ago their best internal model was "somewhat more capable" than Mythos "but does not display a capability jump of the degree observed from Claud

So as of a month ago their best internal model was "somewhat more capable" than Mythos "but does not display a capability jump of the degree observed from Claude Opus 4.6 to Mythos Preview." I thought they would have a significantly more capable model by then, more than five months after Mythos finished training. They'd better have one by now, or the Chinese competitors are closer to catching up than I thought.

andai 2mo ago
You thought they were gonna double the model size again? Also it occurs to me that they're somewhat incentivized to downplay cyber risks after what happened last time...
lumost 2mo ago
I'm still uncertain if mythos is real. Subsequent model releases have been lackluster, no one has claimed to verify mythos performance and it's silently vanished from most comparisons.
- internetter 2mo ago
  Is fable not just mythos with safeguards?
  - lwarfield 2mo ago
    Yes it is: > Same model weights as Mythos 5, deployed with higher-coverage safeguards (see Section 4.5.2.2)