INDEX
Explanations
This neuron detects the discourse marker “Here” used to introduce key methods or findings.
New Auto-Interp
Negative Logits
Maz
-0.07
PST
-0.06
anth
-0.06
istung
-0.06
<
-0.06
Mitarbeiter
-0.06
tits
-0.06
тии
-0.06
sistema
-0.06
.Rect
-0.06
POSITIVE LOGITS
nghiên
0.07
contentPane
0.06
अपन
0.06
[|
0.06
стало
0.06
آینده
0.06
ा↵↵
0.06
DES
0.06
.detectChanges
0.06
welcome
0.06
Activations Density 0.010%