INDEX
Explanations
the word stem "en" that appears at the end of certain words
the presence of the token "en."
New Auto-Interp
Negative Logits
hitch
-0.78
reconc
-0.67
Canaveral
-0.63
pumpkin
-0.63
TPPStreamerBot
-0.63
milo
-0.62
mson
-0.61
psycho
-0.60
Haram
-0.60
shake
-0.59
POSITIVE LOGITS
emies
1.22
chant
1.15
ema
1.07
riched
1.06
viron
1.02
esis
1.01
emy
1.00
zyme
0.99
wald
0.96
igmat
0.96
Activations Density 0.031%