INDEX
Explanations
connections to places and movements
New Auto-Interp
Negative Logits
stro
-0.15
likewise
-0.15
snap
-0.15
ixe
-0.14
kå
-0.14
etch
-0.14
achs
-0.14
ntag
-0.14
Americas
-0.14
дов
-0.14
POSITIVE LOGITS
ge
0.29
ging
0.27
gee
0.26
gesch
0.23
gest
0.23
z
0.22
geh
0.21
zu
0.21
ges
0.21
geb
0.20
Activations Density 0.020%