INDEX
Explanations
mentions of siblings or groups of people with the same last name or title
occurrences of the word "brothers."
New Auto-Interp
Negative Logits
alling
-0.76
okin
-0.69
activity
-0.69
aminer
-0.69
idine
-0.65
orie
-0.65
ORE
-0.64
ointment
-0.64
ainer
-0.63
oval
-0.63
POSITIVE LOGITS
Brothers
0.99
hip
0.96
brothers
0.96
hips
0.95
hood
0.83
fol
0.80
tones
0.77
sisters
0.76
hift
0.75
hes
0.75
Activations Density 0.032%