INDEX
Explanations
occurrences of images in the text
New Auto-Interp
Negative Logits
ambi
-0.19
ække
-0.16
ibs
-0.16
andle
-0.15
amaz
-0.15
aggable
-0.14
ká
-0.14
Bracket
-0.14
retty
-0.14
TL
-0.14
POSITIVE LOGITS
cr
0.16
Ïĥα
0.16
odash
0.15
alto
0.15
igon
0.15
aro
0.14
ÂŃn
0.14
wedge
0.14
vey
0.14
nh
0.14
Activations Density 0.012%