INDEX
Explanations
instances of the word "sometimes."
New Auto-Interp
Negative Logits
somehow
-0.24
ud
-0.17
siempre
-0.16
вÑģегда
-0.16
sempre
-0.15
anyone
-0.15
isser
-0.15
irgend
-0.15
meisten
-0.15
ri
-0.15
POSITIVE LOGITS
even
0.25
-times
0.23
çĶļèĩ³
0.21
even
0.21
referred
0.21
ìĶ©
0.19
侯
0.18
EVEN
0.18
/all
0.18
place
0.18
Activations Density 0.029%