INDEX
Explanations
names of technology companies
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1741
+0.16
0.5%
1984
+0.12
0.4%
1919
+0.11
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1684
+0.16
0.04
1919
+0.12
0.05
619
+0.11
0.04
Negative Logits
Siria
-0.59
dirait
-0.56
parati
-0.56
RectangleBorder
-0.56
EndProject
-0.56
Seeder
-0.55
jaro
-0.55
Referential
-0.54
masaj
-0.54
palab
-0.54
POSITIVE LOGITS
McLaugh
0.71
Wtf
0.63
'
0.63
Whence
0.60
unspeak
0.59
Confu
0.58
’
0.57
wikihow
0.55
withal
0.54
Lmao
0.53
Activations Density 0.246%