INDEX
Explanations
This neuron responds to the surnames of cricket players (i.e. capitalized player last names).
New Auto-Interp
Negative Logits
{@-0.07
得
-0.07
وما
-0.07
imator
-0.06
’ta
-0.06
들을
-0.06
结
-0.06
TextInputType
-0.06
Kürt
-0.06
breakout
-0.06
POSITIVE LOGITS
NSW
0.07
。あ
0.07
нерг
0.06
mentre
0.06
br
0.06
opposing
0.06
leading
0.06
_Row
0.06
підт
0.06
الحي
0.06
Activations Density 0.034%