INDEX
Explanations
conditional phrases and expressions of doubt or contrast
New Auto-Interp
Negative Logits
há
-0.15
Tabs
-0.15
steen
-0.15
iphy
-0.14
ê
-0.14
Tabs
-0.14
.dk
-0.14
etti
-0.14
nesc
-0.14
enheim
-0.14
POSITIVE LOGITS
&
0.16
653
0.16
inf
0.15
ium
0.15
@
0.15
omm
0.15
mention
0.15
imp
0.15
still
0.15
asis
0.15
Activations Density 0.040%