INDEX
Explanations
descriptions of scenes and characters within a setting
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1473
+0.08
0.2%
2010
+0.07
0.2%
382
+0.07
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
2010
+0.08
0.04
1968
+0.07
0.03
1830
+0.07
0.03
Negative Logits
stoff
-0.61
surpl
-0.60
pymysql
-0.59
;;)
-0.55
tille
-0.53
himmel
-0.52
splitContainer
-0.50
trico
-0.50
höl
-0.50
smtplib
-0.50
POSITIVE LOGITS
there
0.71
lies
0.57
there
0.57
theres
0.56
lays
0.54
lüğ
0.52
THERE
0.51
There
0.51
contains
0.48
There
0.48
Activations Density 0.302%