As in, the ratio between (interpretable capabilities / total capabilities) still asymptotes to zero, but the number of interpretable capabilities goes up (and then maybe back down) as the models gain more capabilities?
Yeah. Or maybe not even to zero but it isn’t increasing.
As in, the ratio between (interpretable capabilities / total capabilities) still asymptotes to zero, but the number of interpretable capabilities goes up (and then maybe back down) as the models gain more capabilities?
Yeah. Or maybe not even to zero but it isn’t increasing.