INDEX
Explanations
names of educational institutions along with the word "graduated."
instances of the word "graduated."
New Auto-Interp
Negative Logits
WH
-0.67
BOOK
-0.67
dimension
-0.67
look
-0.67
pro
-0.65
Oo
-0.61
alle
-0.61
erness
-0.60
pg
-0.60
oran
-0.60
POSITIVE LOGITS
uates
1.19
graduated
1.05
uating
0.94
keyes
0.93
uated
0.93
graduates
0.92
uations
0.90
graduating
0.86
uate
0.86
graduation
0.85
Activations Density 0.007%