INDEX
Explanations
instances of the phrase "decided to" indicating choices or actions taken
New Auto-Interp
Negative Logits
reo
-0.16
rière
-0.16
bjerg
-0.15
abyrinth
-0.15
sembler
-0.15
enne
-0.14
riere
-0.14
onym
-0.14
tal
-0.14
urdy
-0.14
POSITIVE LOGITS
atha
0.16
rather
0.16
Wis
0.15
unes
0.14
Alic
0.14
umen
0.13
ometric
0.13
Nay
0.13
616
0.13
hire
0.13
Activations Density 0.025%