INDEX
Explanations
references to Germany and its people
New Auto-Interp
Negative Logits
ciating
-1.03
lihood
-0.83
ording
-0.68
clusively
-0.67
````
-0.66
BOOK
-0.66
ctica
-0.66
Discussion
-0.65
vae
-0.65
clus
-0.65
POSITIVE LOGITS
Chancellor
1.05
wings
1.04
geist
0.97
Bundesliga
0.87
chancellor
0.83
Aerospace
0.81
Bundes
0.80
shire
0.77
Munich
0.76
invaded
0.75
Activations Density 0.021%