INDEX
Explanations
references to geographic or institutional identifiers, specifically "NC" followed by a number
New Auto-Interp
Negative Logits
ugen
-0.18
ugs
-0.15
ono
-0.15
866
-0.15
Berger
-0.15
ington
-0.15
ette
-0.15
descended
-0.15
trp
-0.15
bo
-0.14
POSITIVE LOGITS
abeth
0.17
ến
0.16
sword
0.16
Gear
0.15
rypted
0.14
pls
0.14
Swords
0.14
sage
0.14
sah
0.14
AllWindows
0.13
Activations Density 0.007%