INDEX
Explanations
proper nouns, specifically names
New Auto-Interp
Negative Logits
_VC
-0.16
esk
-0.15
assin
-0.15
bjerg
-0.15
sgi
-0.15
askan
-0.15
acco
-0.15
orde
-0.15
blink
-0.15
oven
-0.15
POSITIVE LOGITS
rick
0.19
trie
0.17
ough
0.17
ory
0.16
vey
0.15
Dough
0.15
aney
0.15
regor
0.15
logue
0.15
cheon
0.15
Activations Density 0.021%