INDEX
Explanations
names and labels followed by specific details
New Auto-Interp
Negative Logits
них
-1.71
cómoda
-1.66
ними
-1.59
mengumumkan
-1.59
delgada
-1.54
metálica
-1.53
they
-1.51
этими
-1.51
garantiza
-1.51
ventajas
-1.50
POSITIVE LOGITS
of
2.61
with
2.33
for
1.85
out
1.77
on
1.73
but
1.66
but
1.52
and
1.42
at
1.41
his
1.39
Activations Density 0.006%