Yea,

I have tested it including

Social reasoning ( which does a good job)

Psychological reasoning ( bad)

Solving math question ( it’s ok, better then Minerva)

Asking LSAT logic game questions ( it gives its thought process, but failed to give correct answers)

I also wrote up a short mystery novel, ( like 200 words, with context) ask if it can tell is the victim is murdered or committed suicide. It actually did ok job on this one if the context is clearly given that everyone can deduce some conclusion using common sense.