INDEX
Explanations
the use of the word "rather" to express preference or contrast
New Auto-Interp
Negative Logits
sf
-0.16
system
-0.15
onders
-0.15
sse
-0.14
sl
-0.14
sm
-0.14
ussen
-0.14
swer
-0.14
imizer
-0.14
ital
-0.14
POSITIVE LOGITS
than
0.37
-than
0.30
than
0.28
_than
0.27
THAN
0.25
Than
0.24
Than
0.22
než
0.22
quam
0.20
än
0.19
Activations Density 0.015%