INDEX
Explanations
mentions of height or tallness
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1350
+0.18
0.7%
1387
+0.16
0.6%
1413
+0.16
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1387
+0.18
0.03
1413
+0.16
0.02
1350
+0.16
0.02
Negative Logits
avoit
-0.58
akku
-0.57
-0.57
perifer
-0.56
fó
-0.55
grossa
-0.55
Sab
-0.53
asfal
-0.52
mahd
-0.51
cár
-0.50
POSITIVE LOGITS
height
1.36
tall
1.29
tall
1.25
height
1.22
Tall
1.22
heights
1.21
Tall
1.19
Height
1.17
Height
1.14
taller
1.12
Activations Density 0.097%