INDEX
Explanations
Croatian language and culture
New Auto-Interp
Negative Logits
şi
2.17
ş
1.89
muß
1.74
ผม
1.72
Ş
1.69
gak
1.66
color
1.60
ผม
1.60
colors
1.59
ţin
1.58
POSITIVE LOGITS
فى
2.23
եւ
1.84
recognisable
1.79
generalised
1.75
التى
1.73
normalised
1.70
Colour
1.65
sólo
1.64
characterised
1.64
minimising
1.62
Activations Density 0.002%