INDEX
Negative Logits
t
1.04
d
0.98
os
0.98
it
0.82
l
0.82
is
0.82
r
0.81
m
0.79
ological
0.76
ac
0.74
POSITIVE LOGITS
DVDs
1.22
DVD
0.96
DVD
0.90
dvd
0.77
шены
0.75
MORDOR
0.74
are
0.71
рина
0.70
VHS
0.70
し
0.70
Activations Density 0.003%
t
d
os
it
l
is
r
m
ological
ac
DVDs
DVD
DVD
dvd
шены
MORDOR
are
рина
VHS
し