INDEX
Explanations
words related to the country of Brazil
references to a specific brand or type of food
New Auto-Interp
Negative Logits
matic
-0.78
Trafford
-0.71
omez
-0.69
aminer
-0.68
aires
-0.67
ARY
-0.67
papers
-0.65
erv
-0.65
ership
-0.63
grounds
-0.62
POSITIVE LOGITS
ided
1.11
ille
1.09
hma
1.09
zz
0.98
very
0.95
VE
0.93
kes
0.93
hm
0.91
ving
0.91
gger
0.90
Activations Density 0.028%