INDEX
Explanations
conjunctions and phrases indicating connection or correlation
New Auto-Interp
Negative Logits
ensibly
-0.16
uC
-0.15
ouv
-0.14
dubious
-0.13
_compat
-0.13
mere
-0.13
other
-0.13
/Internal
-0.13
countless
-0.13
Multiple
-0.13
POSITIVE LOGITS
yet
0.20
full
0.19
yet
0.18
alive
0.17
filled
0.17
border
0.17
verg
0.16
/or
0.16
omba
0.15
alive
0.14
Activations Density 0.315%