INDEX
Explanations
references to the word "Som" and its variations, particularly in relation to Somalia
New Auto-Interp
Negative Logits
ray
-0.18
ξγ
-0.16
تز
-0.16
rompt
-0.16
iert
-0.16
atomic
-0.16
ampoline
-0.15
æ®
-0.15
log
-0.15
edik
-0.15
POSITIVE LOGITS
onto
0.17
lesen
0.17
ewhere
0.17
imbus
0.16
º
0.15
ones
0.15
erville
0.15
cio
0.15
else
0.14
.dispatcher
0.14
Activations Density 0.012%