INDEX
Explanations
descriptive attributes related to objects and their conditions
New Auto-Interp
Negative Logits
estern
-0.72
Ethics
-0.70
Liberals
-0.70
funding
-0.68
policy
-0.67
Liberal
-0.67
Policy
-0.64
Agenda
-0.64
ethical
-0.64
Economic
-0.63
POSITIVE LOGITS
protr
1.32
resemblance
1.08
translucent
1.05
decoration
1.04
rectangular
1.02
decorative
1.02
sculpt
1.00
ornament
0.99
curved
0.98
thickness
0.97
Activations Density 0.437%