INDEX
Explanations
instances of the word "but," indicating contrasting ideas or exceptions
New Auto-Interp
Negative Logits
enberg
-0.19
oha
-0.15
unner
-0.14
aggi
-0.14
pump
-0.14
ãĥ
-0.14
odash
-0.13
oldem
-0.13
ænd
-0.13
gid
-0.13
POSITIVE LOGITS
.semantic
0.18
Ú©ÛĮÙĦ
0.16
ooks
0.16
ifo
0.15
isto
0.14
ION
0.14
WEEN
0.14
ions
0.14
ÏĥÏĥ
0.14
ins
0.13
Activations Density 0.042%