INDEX
Explanations
names of places and people, particularly in a German context
New Auto-Interp
Negative Logits
Auch
-0.15
Yön
-0.14
ancer
-0.14
oksen
-0.14
inch
-0.13
ocab
-0.13
оÑĥ
-0.13
ÑĪиÑĢ
-0.13
ç¦ģ
-0.13
év
-0.13
POSITIVE LOGITS
nat
0.17
stakes
0.16
åľ³
0.15
Germany
0.15
german
0.15
German
0.15
447
0.15
Got
0.14
575
0.14
858
0.14
Activations Density 0.831%