INDEX
Explanations
possessive pronouns and references to personal relationships
New Auto-Interp
Negative Logits
ettel
-0.17
ugo
-0.16
ersist
-0.15
andon
-0.15
eder
-0.15
ichick
-0.15
Kauf
-0.14
ÎŃν
-0.14
alue
-0.14
Cul
-0.14
POSITIVE LOGITS
onward
0.22
wider
0.21
favoured
0.18
allocated
0.17
postcode
0.17
nearest
0.16
calor
0.16
chosen
0.16
advert
0.16
misdemean
0.16
Activations Density 0.188%