INDEX
Explanations
The neuron fires on mentions of the foster‐care system and related terms (e.g. “foster,” “care,” “group-home”).
New Auto-Interp
Negative Logits
ρων
-0.06
Dar
-0.06
dar
-0.06
-Israel
-0.06
ад
-0.06
ッシュ
-0.06
тради
-0.06
:|
-0.06
zap
-0.06
eyin
-0.06
POSITIVE LOGITS
foster
0.09
679
0.07
概
0.07
institutional
0.07
NSMutableArray
0.06
[args
0.06
')),
0.06
overseeing
0.06
obvyk
0.06
LastName
0.06
Activations Density 0.002%