INDEX
Explanations
phrases related to the experiences of being a professional athlete
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1978
+0.11
0.3%
1013
+0.09
0.3%
1654
+0.09
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
658
+0.11
0.05
47
+0.09
0.03
1217
+0.09
0.03
Negative Logits
toscana
-0.72
veneta
-0.58
Personensuche
-0.58
-------------</
-0.56
<<<<<<<<<<<<<<
-0.56
utop
-0.55
algas
-0.53
pittores
-0.53
fré
-0.53
poros
-0.52
POSITIVE LOGITS
blackpink
0.75
hairc
0.74
conftest
0.73
Shakspeare
0.72
ecru
0.71
swarovski
0.70
timately
0.68
tucson
0.67
wikihow
0.67
churrasco
0.67
Activations Density 0.342%