INDEX
Explanations
conjunctions indicating contrast or alternative options
the conjunction "but" in contexts contrasting different ideas or conditions
New Auto-Interp
Negative Logits
acca
-0.67
onne
-0.62
amus
-0.62
leen
-0.60
lein
-0.59
LU
-0.59
rill
-0.59
lement
-0.59
itia
-0.58
atoon
-0.57
POSITIVE LOGITS
none
0.96
alas
0.83
none
0.76
unfortunately
0.76
tons
0.72
nil
0.71
nothing
0.71
ignored
0.70
declined
0.69
mostly
0.69
Activations Density 0.430%