INDEX
Explanations
instances of the word "brilliant."
New Auto-Interp
Negative Logits
-0.17
oples
-0.15
olland
-0.14
strcasecmp
-0.14
blind
-0.14
/people
-0.14
份
-0.14
zman
-0.14
Religion
-0.13
/licenses
-0.13
POSITIVE LOGITS
readcr
0.15
ong
0.15
/*@
0.14
adu
0.14
æk
0.14
/ext
0.14
ummer
0.13
tsky
0.13
mente
0.13
rams
0.13
Activations Density 0.005%