INDEX
Explanations
This neuron responds to occurrences of the word “carrier.”
New Auto-Interp
Negative Logits
"How
-0.07
pregunta
-0.06
-degree
-0.06
redits
-0.06
Sug
-0.06
ضو
-0.06
Ud
-0.06
Nicola
-0.06
experiments
-0.06
unnable
-0.06
POSITIVE LOGITS
carrier
0.11
Carrier
0.11
Carrier
0.09
carriers
0.09
er
0.08
carrier
0.08
corresponding
0.08
ר
0.08
es
0.07
-ar
0.07
Activations Density 0.003%