Standard level of 2023-2024 intelligence for a 4-8B
#2
by
BingoBird - opened
Generally this class of model lacks the ability to infer things by applying un-mentioned knowledge to a given problem.
Out-of-sight, out-of-mind unfortunately applies to this one, in my tests.
Maybe good for "shape some prose along these lines", not "this character needs to solve this dilemma" problems.
I'm sorry I can't provide a review for this for the intended roleplay/prose use.
It's good?