INDEX
Explanations
numerical representations and measurements related to growth or size
New Auto-Interp
Head Attr Weights
0:0.02
1:0.02
2:0.12
3:0.11
4:0.20
5:0.03
6:0.04
7:0.16
8:0.04
9:0.06
10:0.08
11:0.06
Negative Logits
consulted
-1.74
WRITE
-1.68
wrote
-1.66
listened
-1.57
ighting
-1.54
ocus
-1.50
editors
-1.47
]]
-1.45
communicated
-1.44
rette
-1.44
POSITIVE LOGITS
lengths
1.94
reservoirs
1.88
Balt
1.74
extremes
1.73
ldom
1.72
manageable
1.67
actly
1.64
peaks
1.62
ishable
1.57
tle
1.56
Activations Density 0.000%