INDEX
Explanations
phrases related to facial features or characteristics
terms related to facial recognition or facial features
New Auto-Interp
Negative Logits
mate
-0.68
breath
-0.65
Seas
-0.64
grandson
-0.64
grandfather
-0.63
die
-0.63
brother
-0.63
undone
-0.62
hood
-0.62
Wink
-0.62
POSITIVE LOGITS
ilitation
1.34
ulty
1.34
eless
1.26
etime
1.24
ep
1.16
ilitating
1.15
ilit
1.13
sim
1.10
ework
1.08
ility
1.04
Activations Density 0.070%