INDEX
Explanations
films, immediately
This neuron detects HTML/JSX form markup and input-related attributes (i.e., code tokens tied to form fields and their properties).
New Auto-Interp
Negative Logits
인정
0.50
形で
0.45
vorgesehen
0.45
世紀
0.44
Meteorology
0.44
ಓ
0.43
ResourceManager
0.43
<0x06>
0.41
हैद
0.41
പ്പിക്കുന്ന
0.40
POSITIVE LOGITS
flapping
0.43
er
0.43
ak
0.42
fluid
0.42
Sa
0.42
flipped
0.42
u
0.41
above
0.40
org
0.40
上面
0.39
Activations Density 0.216%