INDEX
Explanations
references to fictional or mythical characters and entities
references to "beard" and related terms, as well as contexts involving a lookout or vigil
New Auto-Interp
Negative Logits
icken
-0.73
arcity
-0.69
Pradesh
-0.64
alls
-0.64
psychic
-0.61
tones
-0.61
raints
-0.60
scam
-0.60
lap
-0.57
cause
-0.56
POSITIVE LOGITS
beard
1.45
Deploy
0.71
ufact
0.68
romeda
0.67
ching
0.66
lished
0.65
redes
0.64
chwitz
0.63
sacrific
0.63
llor
0.62
Activations Density 0.001%