INDEX
Explanations
variations of the word "form" in different contexts
New Auto-Interp
Negative Logits
ilt
-0.16
forefront
-0.15
opus
-0.15
ra
-0.14
ours
-0.14
wyn
-0.14
dek
-0.14
emia
-0.14
counter
-0.14
aurus
-0.14
POSITIVE LOGITS
ulating
0.17
idable
0.16
/form
0.15
ostel
0.15
unately
0.15
teenth
0.14
ulary
0.14
indeb
0.14
(forms
0.14
PerPixel
0.14
Activations Density 0.039%