INDEX
Explanations
references to ears and ear-related items or concepts
New Auto-Interp
Negative Logits
enko
-0.21
eur
-0.18
avor
-0.17
dorf
-0.17
eer
-0.16
yah
-0.15
datagrid
-0.15
erson
-0.15
Pont
-0.15
.strict
-0.15
POSITIVE LOGITS
phones
0.31
phone
0.28
lobe
0.26
bud
0.26
then
0.25
liest
0.24
lob
0.24
piece
0.23
ache
0.23
shot
0.22
Activations Density 0.010%