INDEX
Explanations
names, especially with the name "Bernard", and also potentially legal or academic terms
mentions of specific individuals, particularly those named Bernard and Allison
New Auto-Interp
Negative Logits
pter
-0.99
onductor
-0.93
fulness
-0.84
eah
-0.78
pered
-0.78
med
-0.76
ord
-0.75
appers
-0.74
mitted
-0.74
asio
-0.73
POSITIVE LOGITS
naire
0.94
rage
0.83
ous
0.82
hof
0.80
âĸĪ
0.77
jee
0.73
ial
0.73
ged
0.71
naires
0.70
ger
0.67
Activations Density 0.153%