INDEX
Explanations
statistics and data related to various topics
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1343
+0.13
0.4%
609
+0.10
0.3%
1013
+0.09
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
609
+0.13
0.04
1559
+0.10
0.03
1435
+0.09
0.04
Negative Logits
shenan
-0.95
melange
-0.89
sophistic
-0.87
unspeak
-0.85
pamph
-0.79
languid
-0.79
rascal
-0.79
frivol
-0.79
poetical
-0.78
conceit
-0.78
POSITIVE LOGITS
alkoh
1.26
silikon
1.18
kristal
1.18
karton
1.16
maksi
1.04
uhr
1.03
kosme
1.01
augus
1.01
tomat
1.01
seksi
1.00
Activations Density 0.083%