INDEX
Explanations
references to the concept of transcendence
New Auto-Interp
Negative Logits
trag
-0.15
isque
-0.15
steen
-0.15
léd
-0.15
embro
-0.15
tery
-0.14
tings
-0.14
tlement
-0.14
istry
-0.14
esseract
-0.14
POSITIVE LOGITS
endent
0.29
ental
0.26
enders
0.22
ending
0.21
ende
0.21
endance
0.20
ence
0.20
end
0.20
ended
0.19
ients
0.19
Activations Density 0.005%