INDEX
Explanations
references to readers and their engagement with the text
New Auto-Interp
Negative Logits
Ceinture
-0.80
irinha
-0.78
DockStyle
-0.78
wnia
-0.78
før
-0.75
Commandant
-0.74
PerformLayout
-0.73
Skinny
-0.71
Olímpicos
-0.71
InjectAttribute
-0.71
POSITIVE LOGITS
Readers
1.53
readers
1.51
Reader
1.43
reader
1.35
readers
1.33
Readers
1.33
Reader
1.24
reader
1.23
READER
1.17
audience
1.11
Activations Density 0.062%