INDEX
Negative Logits
kloped
-0.81
houſe
-0.78
iſt
-0.76
noastre
-0.75
pleaſure
-0.75
Houſe
-0.74
Roskov
-0.74
envolvimento
-0.73
becauſe
-0.72
itſelf
-0.72
POSITIVE LOGITS
many
0.54
scores
0.54
multiple
0.54
several
0.52
mark
0.52
twice
0.52
numerous
0.51
0.49
con
0.48
three
0.47
Activations Density 0.158%