INDEX
Explanations
references to bathrooms and related features in a property description
New Auto-Interp
Negative Logits
ito
-0.17
idel
-0.17
triang
-0.15
asil
-0.15
subpoena
-0.15
542
-0.15
uforia
-0.15
872
-0.15
igham
-0.14
ırak
-0.14
POSITIVE LOGITS
rete
0.18
afil
0.16
agra
0.15
anst
0.15
enze
0.15
kup
0.15
Ñĥг
0.15
adesh
0.15
.cg
0.14
Robbins
0.14
Activations Density 0.005%