INDEX
Explanations
terms related to conversion, particularly in the context of religion and identity
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1328
+0.14
0.5%
90
+0.14
0.5%
331
+0.11
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1328
+0.14
0.02
569
+0.14
0.02
90
+0.11
0.02
Negative Logits
Estrella
-0.51
Bahía
-0.49
sembrano
-0.48
abbiano
-0.48
parteci
-0.47
terea
-0.47
dovrebbero
-0.47
czegó
-0.46
Obrázky
-0.45
sacerd
-0.45
POSITIVE LOGITS
Conversion
1.20
conversion
1.18
conversions
1.18
Conversions
1.18
Conversion
1.10
converts
1.10
convert
1.09
conversion
1.06
converter
1.04
Converts
1.03
Activations Density 0.063%