INDEX
Explanations
text related to programming, particularly involving strings, methods, and function calls
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
2019
+0.18
0.6%
453
+0.15
0.5%
876
+0.14
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1404
+0.18
0.03
545
+0.15
0.03
453
+0.14
0.03
Negative Logits
,
-0.76
as
-0.73
le
-0.71
per
-0.70
.
-0.70
so
-0.70
for
-0.69
he
-0.69
-0.69
but
-0.68
POSITIVE LOGITS
sappi
1.59
fordable
1.52
dises
1.47
bourgeo
1.41
embra
1.40
timately
1.39
viciss
1.38
embodi
1.36
simplif
1.36
guatemala
1.36
Activations Density 0.093%