INDEX
Explanations
references to familial relationships and ancestry
New Auto-Interp
Negative Logits
466
-0.16
ebek
-0.15
符
-0.15
oined
-0.15
lyph
-0.14
oard
-0.14
oleans
-0.14
APH
-0.14
Codec
-0.14
pdb
-0.13
POSITIVE LOGITS
distant
0.33
kin
0.29
related
0.28
relation
0.25
Related
0.25
relations
0.24
cousins
0.24
Close
0.24
related
0.23
descend
0.23
Activations Density 0.162%