INDEX
Explanations
phrases related to the development and details of a particular product or project
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1013
+0.09
0.3%
1056
+0.09
0.2%
1703
+0.08
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1056
+0.09
0.04
194
+0.09
0.01
714
+0.08
0.04
Negative Logits
Realt
-0.79
ecru
-0.78
cushi
-0.76
cabrio
-0.73
vespa
-0.73
lpg
-0.72
OGS
-0.72
PLW
-0.71
oleo
-0.71
trefoil
-0.71
POSITIVE LOGITS
<bos>
0.88
inspiration
0.67
inspired
0.65
noticing
0.58
Inspired
0.56
brainstorming
0.55
Inspired
0.55
İstinadlar
0.54
Conheça
0.54
noticed
0.53
Activations Density 0.452%