INDEX
Explanations
mentions of Germany or German-related entities
New Auto-Interp
Negative Logits
ciating
-0.95
lihood
-0.75
ctica
-0.74
ording
-0.70
````
-0.70
heed
-0.69
vae
-0.68
clus
-0.66
clusively
-0.66
BOOK
-0.66
POSITIVE LOGITS
Chancellor
1.12
wings
1.12
geist
0.94
Bundesliga
0.94
chancellor
0.89
Bundes
0.85
shire
0.81
shepherd
0.81
Munich
0.80
Aerospace
0.80
Activations Density 0.036%