INDEX
Explanations
information related to technical instructions and troubleshooting for devices such as printers, generators, and computers
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.18
0.8%
1385
+0.10
0.5%
1438
+0.10
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1385
+0.18
0.10
1819
+0.10
0.07
658
+0.10
0.07
Negative Logits
<bos>
-3.14
intersper
-0.86
endow
-0.80
disbur
-0.79
abate
-0.77
repress
-0.76
promul
-0.76
amass
-0.75
reconno
-0.72
cede
-0.71
POSITIVE LOGITS
">...
0.95
jawa
0.85
">/
0.84
ecru
0.79
bandung
0.77
romero
0.76
>\<
0.73
Bekasi
0.73
quoc
0.72
Lampung
0.72
Activations Density 0.598%