INDEX
Explanations
references to familial relationships and lineage
New Auto-Interp
Negative Logits
Uns
-0.18
uche
-0.17
collo
-0.15
oje
-0.14
uns
-0.14
UCT
-0.14
uchen
-0.14
lator
-0.14
Tort
-0.14
nost
-0.13
POSITIVE LOGITS
Dims
0.16
Hdr
0.15
igt
0.15
chrom
0.15
Butler
0.14
701
0.14
ëı
0.14
иг
0.14
.BL
0.14
ig
0.14
Activations Density 0.070%