INDEX
Explanations
phrases related to the political sphere or government actions
sentences ending in a period
New Auto-Interp
Negative Logits
mosqu
-0.74
iber
-0.67
culus
-0.67
isphere
-0.67
uly
-0.66
defe
-0.64
eme
-0.64
purse
-0.63
thal
-0.63
ascus
-0.62
POSITIVE LOGITS
However
1.12
Specifically
1.08
Especially
1.06
Whether
1.04
They
1.04
Though
1.03
While
1.01
But
1.01
Unfortunately
1.00
Although
0.99
Activations Density 1.202%