INDEX
Explanations
keywords related to government, media, and economy
New Auto-Interp
Head Attr Weights
0:0.35
1:0.04
2:0.01
3:0.07
4:0.11
5:0.07
6:0.04
7:0.04
8:0.07
9:0.09
10:0.02
11:0.03
Negative Logits
ゴン
-1.65
specific
-1.64
wide
-1.62
Spot
-1.59
inite
-1.51
Squid
-1.50
atively
-1.50
aneous
-1.47
phosphate
-1.46
outheast
-1.45
POSITIVE LOGITS
christ
1.66
ruining
1.60
headlined
1.60
fuck
1.59
#$
1.58
whose
1.56
flanked
1.54
purse
1.51
hauled
1.50
!!!!
1.49
Activations Density 0.001%