INDEX
Explanations
elements describing characters and relationships in narratives
New Auto-Interp
Negative Logits
takson
-0.16
ystack
-0.16
achen
-0.15
olume
-0.14
orer
-0.14
apiro
-0.14
æ°ı
-0.14
readcr
-0.14
ergarten
-0.14
='../
-0.13
POSITIVE LOGITS
inher
0.15
otal
0.15
ÙĪÙĬÙĥ
0.14
SCRI
0.14
dom
0.14
conc
0.14
WikiLeaks
0.14
ÙĬÙĥÙĬ
0.13
ego
0.13
èŃ
0.13
Activations Density 0.028%