INDEX
Explanations
numerical data and values
New Auto-Interp
Negative Logits
oya
-0.16
583
-0.14
į¨
-0.13
/target
-0.13
Eth
-0.13
nob
-0.13
enburg
-0.13
(Target
-0.13
ifer
-0.13
imes
-0.13
POSITIVE LOGITS
ysa
0.17
rieg
0.15
tü
0.15
ohana
0.15
вано
0.14
iyim
0.14
pus
0.14
vore
0.14
dit
0.14
orr
0.14
Activations Density 0.011%