INDEX
Explanations
references to bodily sensations and physical conditions
New Auto-Interp
Negative Logits
θι
-0.15
inval
-0.15
éc
-0.14
erap
-0.14
hip
-0.14
head
-0.14
Voc
-0.14
彡
-0.14
Sharp
-0.13
bos
-0.13
POSITIVE LOGITS
nat
0.17
ottage
0.17
.react
0.16
uis
0.15
éħ
0.14
/body
0.14
bud
0.14
esz
0.14
.function
0.14
zeigen
0.14
Activations Density 0.145%