INDEX
Explanations
information related to political events and geographical locations
New Auto-Interp
Negative Logits
wa
-0.68
vic
-0.68
tailed
-0.65
illed
-0.63
ashes
-0.62
anut
-0.61
uese
-0.61
achus
-0.60
alsa
-0.59
tex
-0.59
POSITIVE LOGITS
resembles
0.82
resemble
0.75
resembled
0.71
parallels
0.68
analogous
0.67
mirror
0.66
terday
0.66
mite
0.65
paralle
0.61
mirrors
0.61
Activations Density 0.045%