INDEX
Explanations
mentions of the name "Graham" at various points
New Auto-Interp
Negative Logits
puter
-0.93
merce
-0.85
senal
-0.77
————————
-0.76
cles
-0.74
selves
-0.71
Rated
-0.69
worldly
-0.68
ngth
-0.66
haps
-0.66
POSITIVE LOGITS
esse
0.85
Greene
0.83
Graham
0.82
iday
0.76
iar
0.74
Ń·
0.74
sburg
0.74
wagen
0.72
Hancock
0.71
etti
0.68
Activations Density 0.100%