INDEX
Explanations
phrases related to activities or events happening at night
mentions of "night."
New Auto-Interp
Negative Logits
achev
-0.75
VE
-0.69
ANY
-0.68
berman
-0.67
Æ
-0.67
bearing
-0.65
agles
-0.64
ences
-0.63
FW
-0.63
ENCE
-0.63
POSITIVE LOGITS
mares
1.44
night
1.29
mare
1.17
urnal
0.99
Night
0.89
bird
0.81
endment
0.80
theless
0.78
loo
0.77
lihood
0.76
Activations Density 0.007%