INDEX
Explanations
words related to a specific individual, possibly a celebrity
the name "Joko" and its variations
New Auto-Interp
Negative Logits
bred
-0.77
chall
-0.69
Colorado
-0.65
spin
-0.64
docs
-0.64
rained
-0.63
ATTLE
-0.62
weeney
-0.62
pent
-0.62
models
-0.62
POSITIVE LOGITS
Haram
1.31
oko
1.27
wana
0.99
ichi
0.96
ji
0.93
Kus
0.88
iah
0.88
ibo
0.87
oji
0.80
ovy
0.79
Activations Density 0.007%