INDEX
Explanations
terms related to housing, such as "housing market," "housing development," and "public flats."
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.20
1.2%
1472
+0.10
0.6%
1406
+0.10
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1406
+0.20
0.03
486
+0.10
0.02
976
+0.10
0.03
Negative Logits
<bos>
-3.16
<?
-1.19
-1.07
ⓧ
-1.04
/**
-0.94
/***
-0.89
<?
-0.77
/*
-0.71
defray
-0.65
inaugurate
-0.63
POSITIVE LOGITS
Housing
1.11
housing
1.10
Housing
1.08
maroc
1.06
HOUSING
0.99
Minang
0.97
housing
0.95
dises
0.94
saar
0.93
Muhamma
0.92
Activations Density 0.104%