INDEX
Explanations
available
This neuron activates chiefly on the word “available” (and closely related tokens indicating availability).
New Auto-Interp
Negative Logits
Converted
-0.07
Jacob
-0.07
Mahon
-0.06
Дмит
-0.06
Adolf
-0.06
Phill
-0.06
Typ
-0.06
Αγγ
-0.06
مبت
-0.06
Chase
-0.06
POSITIVE LOGITS
hoá
0.07
$field
0.07
させる
0.06
seeing
0.06
'name
0.06
_buttons
0.06
setLabel
0.06
remembering
0.06
<Unit
0.06
////////////////////////////////////////////////////////////////////////////////
0.06
Activations Density 0.035%