INDEX
Explanations
the presence of the suffix "el" or similar phonetic endings in words
New Auto-Interp
Negative Logits
lian
-0.18
lopedia
-0.18
ibble
-0.18
enze
-0.18
estre
-0.15
ãĤ´ãĥª
-0.15
egl
-0.15
baz
-0.15
led
-0.14
rale
-0.14
POSITIVE LOGITS
cano
0.16
bv
0.15
subsid
0.15
idon
0.15
alem
0.15
ldr
0.15
xn
0.14
assi
0.14
subsidized
0.14
cir
0.14
Activations Density 0.027%