INDEX
Explanations
words related to natural phenomena and specific times of day
references to sunsets and sunrises
New Auto-Interp
Negative Logits
onne
-0.85
ollow
-0.80
ongh
-0.80
urers
-0.79
artments
-0.78
ioch
-0.77
iott
-0.75
oyal
-0.72
enne
-0.72
erto
-0.72
POSITIVE LOGITS
sunset
0.94
sunrise
0.87
Sunrise
0.85
Sunset
0.84
shine
0.82
issance
0.82
Sparkle
0.79
creen
0.76
slit
0.76
lit
0.75
Activations Density 0.011%