INDEX
Negative Logits
-0.08
adne
-0.07
spent
-0.07
knob
-0.07
bucket
-0.07
pointer
-0.07
wager
-0.07
meter
-0.07
Papa
-0.07
(bucket
-0.07
POSITIVE LOGITS
successor
0.09
successors
0.09
amiz
0.08
’y
0.08
паш
0.08
amist
0.08
reempl
0.08
рол
0.08
enary
0.08
ня
0.07
Activations Density 0.011%