INDEX
Explanations
mentions of the state of Michigan in news articles
references to the state of Michigan
New Auto-Interp
Negative Logits
ciating
-0.84
istics
-0.83
bable
-0.80
footed
-0.79
balance
-0.78
fty
-0.73
figure
-0.71
desks
-0.70
ngth
-0.69
hedral
-0.68
POSITIVE LOGITS
Mich
1.18
Mich
1.03
onne
0.90
alis
0.89
ér
0.83
olt
0.81
ol
0.81
omed
0.80
ias
0.76
olit
0.76
Activations Density 0.005%