INDEX
Explanations
references to "knock" or "knock-off" related words or phrases
New Auto-Interp
Negative Logits
uzey
-0.18
ÑĢава
-0.17
imbus
-0.16
uite
-0.15
alyzed
-0.15
ɵ
-0.14
_finalize
-0.14
prus
-0.14
igned
-0.14
needle
-0.13
POSITIVE LOGITS
knock
0.26
Knock
0.25
down
0.25
Kn
0.23
kn
0.22
offs
0.22
-k
0.21
knocks
0.21
knocking
0.20
knocked
0.19
Activations Density 0.005%