INDEX
Explanations
mentions of geological features and characteristics
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1577
+0.51
2.0%
50
+0.22
0.8%
381
+0.11
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1577
+0.51
0.20
184
+0.22
0.07
1965
+0.11
0.10
Negative Logits
mercen
-0.55
churrasco
-0.52
Darío
-0.49
Áng
-0.47
Mónica
-0.47
curé
-0.46
membrance
-0.46
tortas
-0.46
paillettes
-0.45
calciatore
-0.45
POSITIVE LOGITS
inoltre
0.65
furthermore
0.50
Expt
0.47
however
0.46
aen
0.46
moreover
0.45
!'
0.45
:,,
0.44
also
0.44
ouncil
0.44
Activations Density 3.871%