INDEX
Explanations
dates and numbers
The neuron detects date/time tokens and explicit temporal references (months, days, years, and timestamps).
New Auto-Interp
Negative Logits
arrhythmias
0.33
Над
0.29
蜢
0.29
δες
0.29
последствия
0.29
Пре
0.28
ﺍﻟ
0.28
吣
0.28
Вер
0.27
Ш
0.27
POSITIVE LOGITS
の
0.44
or
0.38
2
0.37
F
0.37
.
0.35
nya
0.34
ay
0.33
os
0.33
un
0.30
G
0.30
Activations Density 0.201%