INDEX
Explanations
aspects related to family and social relationships
Third-person pronouns followed by auxiliary verbs
did not, cannot, would not
New Auto-Interp
Negative Logits
weren
-1.18
aren
-1.17
isn
-1.10
couldn
-1.09
couldn
-1.05
wasn
-1.04
shouldn
-1.02
nggak
-0.99
hasn
-0.99
wasn
-0.97
POSITIVE LOGITS
cannot
0.88
cannot
0.81
Cannot
0.79
Cannot
0.70
!!!!!!!!!!!!!!!!
0.65
のである
0.64
expandindo
0.63
does
0.57
"><?=
0.56
did
0.55
Activations Density 0.460%