INDEX
Explanations
comparisons between original designs and their replicas or imitations
New Auto-Interp
Negative Logits
GRESS
-0.16
ODE
-0.16
umen
-0.15
Crest
-0.14
Cust
-0.14
Bett
-0.14
277
-0.14
139
-0.14
Trang
-0.13
opro
-0.13
POSITIVE LOGITS
originals
0.37
original
0.30
original
0.26
-original
0.25
оÑĢиг
0.25
ORIGINAL
0.23
/original
0.22
actual
0.22
Original
0.22
_original
0.22
Activations Density 0.132%