INDEX
Explanations
percentages or numerical values denoting an increase or a change over time
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.20
0.8%
143
+0.07
0.3%
2000
+0.07
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
143
+0.20
0.04
378
+0.07
0.03
2000
+0.07
0.04
Negative Logits
<bos>
-2.32
ⓧ
-1.01
-0.89
/**
-0.79
afterEach
-0.73
<?
-0.70
/***
-0.70
///**
-0.65
/*!
-0.64
<?
-0.61
POSITIVE LOGITS
swarovski
1.25
ecru
1.24
kawasaki
1.24
maneu
1.23
bandung
1.19
lamborghini
1.19
chrysler
1.18
eiffel
1.17
mitsubishi
1.17
jbl
1.13
Activations Density 0.274%