INDEX
Explanations
Medical texts
The neuron strongly activates on numeric expressions and related statistical or measurement tokens (e.g. digits, percentages, p-values, and other quantitative data).
New Auto-Interp
Negative Logits
plans
-0.07
dispute
-0.07
newsletter
-0.07
Plans
-0.06
＿＿＿＿
-0.06
converted
-0.06
odi
-0.06
.units
-0.06
услов
-0.06
ارائه
-0.06
POSITIVE LOGITS
uckles
0.07
إد
0.06
kapas
0.06
villa
0.06
ъем
0.06
Cheney
0.06
syll
0.06
Ori
0.06
рем
0.06
�
0.06
Activations Density 0.083%