INDEX
Explanations
people's names, particularly the surname "Richards"
references to the name "Richards."
New Auto-Interp
Negative Logits
phis
-0.94
anooga
-0.83
unal
-0.75
phas
-0.71
sight
-0.67
pneum
-0.66
dolphin
-0.65
ocular
-0.65
alid
-0.64
oxin
-0.63
POSITIVE LOGITS
Richards
1.06
mond
0.86
lings
0.80
zman
0.79
cream
0.79
burgh
0.78
bard
0.70
het
0.70
heim
0.69
hips
0.68
Activations Density 0.003%