INDEX
Explanations
the repetition of the token "Du."
New Auto-Interp
Negative Logits
getLogger
-0.66
Gund
-0.60
tapan
-0.59
syke
-0.59
zás
-0.57
ποίη
-0.55
życie
-0.54
Vickers
-0.53
ίων
-0.53
commitments
-0.53
POSITIVE LOGITS
Du
1.37
Du
1.27
du
1.25
DU
1.05
du
1.03
DU
0.94
thou
0.91
Thou
0.84
Dubois
0.75
ду
0.74
Activations Density 0.096%