INDEX
Explanations
version numbers in software-related texts
New Auto-Interp
Negative Logits
383
-0.18
enÃŃ
-0.16
erial
-0.15
onto
-0.15
aan
-0.14
aks
-0.14
htar
-0.14
836
-0.14
Steele
-0.13
ioni
-0.13
POSITIVE LOGITS
azen
0.18
ziel
0.16
ket
0.16
ascript
0.15
bish
0.15
ulur
0.15
native
0.14
nech
0.14
Invocation
0.14
ãĥIJãĥ¼
0.14
Activations Density 0.029%