INDEX
Explanations
phrases mentioning the weekend
references to weekends
New Auto-Interp
Negative Logits
metics
-0.90
eus
-0.83
ymes
-0.78
abet
-0.78
rator
-0.77
ancy
-0.77
itivity
-0.75
itative
-0.75
mitter
-0.75
ifice
-0.74
POSITIVE LOGITS
mornings
1.18
afternoon
1.08
nights
1.01
morning
0.97
evening
0.94
night
0.94
night
0.93
weekend
0.93
evenings
0.93
brunch
0.86
Activations Density 0.017%