INDEX
Explanations
references to shapes and forms that are round or curved
New Auto-Interp
Negative Logits
æĵ¦
-0.16
apult
-0.16
é¡Ķ
-0.15
پرد
-0.15
ivot
-0.14
nesia
-0.14
.ua
-0.14
æľŃ
-0.14
каÑģ
-0.14
---</
-0.14
POSITIVE LOGITS
atters
0.18
0.17
y
0.17
ohl
0.15
nature
0.15
mere
0.14
Form
0.14
,
0.14
yd
0.14
ery
0.14
Activations Density 0.043%