INDEX
Explanations
references to players in sports contexts
New Auto-Interp
Negative Logits
ertype
-0.21
ibility
-0.18
resse
-0.17
sWith
-0.17
arget
-0.16
ensi
-0.16
igkeit
-0.16
/windows
-0.15
ral
-0.15
inue
-0.15
POSITIVE LOGITS
hood
0.26
hip
0.25
who
0.24
innen
0.24
hips
0.24
unions
0.23
union
0.20
whom
0.20
Who
0.19
/client
0.19
Activations Density 0.105%