INDEX
Explanations
punctuation
The neuron fires on direct imperative commands—especially instructions to undress or assume sexual/punishment positions (e.g. “take off your clothes,” “get on all fours”).
New Auto-Interp
Negative Logits
isas
-0.07
Гол
-0.06
خوب
-0.06
النس
-0.06
şk
-0.06
acoes
-0.06
如此
-0.06
lenir
-0.06
/forum
-0.06
infinit
-0.06
POSITIVE LOGITS
erf
0.07
Bj
0.07
顶
0.07
(**
0.07
(helper
0.07
conna
0.06
pys
0.06
(Container
0.06
%">↵
0.06
код
0.06
Activations Density 0.057%