INDEX
Explanations
the word "do" and related phrases in the context of carrying out tasks or activities
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
90
+0.15
0.5%
370
+0.12
0.4%
878
+0.10
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
90
+0.15
0.08
370
+0.12
0.07
1811
+0.10
0.06
Negative Logits
RefreshLayout
-0.51
΄
-0.51
AreEqual
-0.49
transQ
-0.49
ArgumentParser
-0.48
odkazy
-0.48
Composable
-0.47
ves
-0.47
menta
-0.47
Partici
-0.47
POSITIVE LOGITS
fta
1.29
encomp
1.24
intersper
1.19
squa
1.18
increa
1.16
disagre
1.16
reluct
1.15
thut
1.13
swarovski
1.13
ftu
1.13
Activations Density 0.155%