INDEX
Explanations
names related to politics
references to the individual named Chaffetz
New Auto-Interp
Negative Logits
ocene
-0.75
Clockwork
-0.69
————————————————
-0.66
condition
-0.65
³³³³³³³³³³³³³³³³
-0.63
ynthesis
-0.60
STD
-0.60
Lisbon
-0.60
WARD
-0.59
Archdemon
-0.58
POSITIVE LOGITS
Chaff
1.19
etz
1.18
Emin
0.89
eneg
0.89
bats
0.82
inary
0.82
rons
0.81
ey
0.78
rey
0.78
illo
0.78
Activations Density 0.009%