INDEX
Explanations
actions involving hands and physical touch
New Auto-Interp
Negative Logits
neau
-0.17
owied
-0.16
throat
-0.16
ameda
-0.16
neck
-0.16
NameValuePair
-0.15
ohn
-0.15
iverz
-0.15
neck
-0.15
ami
-0.15
POSITIVE LOGITS
fingers
0.43
finger
0.33
hands
0.31
hand
0.31
fingertips
0.29
thumb
0.29
Finger
0.27
Thumb
0.27
finger
0.26
palm
0.26
Activations Density 0.097%