INDEX
Explanations
\textbf{Now, I'm sorry, but it seems like the neuron's activations provided are too fragmented, and it's unsure what the neuron is looking for. It would be best if you reexamine the neuron's activations, and then I'll be glad to assist you in understanding what
the character 'âĢ' with varying frequencies
New Auto-Interp
Negative Logits
scatter
-0.72
anwhile
-0.72
shack
-0.70
dividing
-0.69
mileage
-0.69
econom
-0.64
fixture
-0.62
infringing
-0.62
tremend
-0.62
measuring
-0.62
POSITIVE LOGITS
¬
1.04
¡
1.02
º
1.02
Ń
1.01
ı
1.00
¹
0.99
²
0.99
į
0.99
Ī
0.99
Į
0.96
Activations Density 0.187%