INDEX
Explanations
attends to sentiments related to being fortunate or lucky from tokens discussing birth or existence
New Auto-Interp
Head Attr Weights
0:0.15
1:0.22
2:0.14
3:0.06
4:0.05
5:0.02
6:0.05
7:0.27
Negative Logits
мәкал
-0.42
NameInMap
-0.39
adaptiveStyles
-0.36
tvguidetime
-0.33
vician
-0.33
MemoryWarning
-0.32
ValueStyle
-0.32
Personensuche
-0.32
autorytatywna
-0.32
存于互联网档案馆
-0.32
POSITIVE LOGITS
stringBuilder
0.26
besch
0.25
Viited
0.23
Kanpo
0.23
elling
0.22
devtools
0.22
);
0.21
りましたが
0.21
plar
0.20
Тип
0.20
Activations Density 0.083%