INDEX
Explanations
references to disparities in health care and education among demographic groups
New Auto-Interp
Negative Logits
unal
-0.16
apl
-0.15
Åŀah
-0.15
ugins
-0.14
aber
-0.14
Dickinson
-0.14
Ñħи
-0.14
elve
-0.14
롱
-0.14
tridge
-0.14
POSITIVE LOGITS
likelihood
0.32
lik
0.30
more
0.29
tend
0.28
likely
0.28
twice
0.27
likelihood
0.27
unlikely
0.24
tends
0.24
Likely
0.23
Activations Density 0.099%