INDEX
Explanations
This neuron specifically detects incest-related content tags (e.g. “incest,” “brother/sister”) in the prompt’s metadata.
New Auto-Interp
Negative Logits
Ramos
-0.08
twig
-0.08
Ling
-0.08
(resultSet
-0.06
Mart
-0.06
(reg
-0.06
())
-0.06
,response
-0.06
<Guid
-0.06
-0.06
POSITIVE LOGITS
_GREEN
0.06
tvor
0.06
Changes
0.06
sí
0.06
usher
0.06
_visitor
0.06
ifty
0.06
initWithTitle
0.06
.Expressions
0.06
.forms
0.06
Activations Density 0.009%