INDEX
Explanations
references to people passing away or dying
phrases related to death and passing away
New Auto-Interp
Negative Logits
Pros
-0.75
rouse
-0.71
oS
-0.67
Ŀ
-0.65
Choice
-0.65
uu
-0.65
amac
-0.65
inge
-0.64
escal
-0.64
ustomed
-0.63
POSITIVE LOGITS
tragically
0.85
psychiat
0.83
retire
0.81
thood
0.77
prematurely
0.73
deceased
0.73
retiring
0.71
peacefully
0.70
intest
0.70
retirement
0.69
Activations Density 0.172%