INDEX
Explanations
references to a specific book or concept (The Secret)
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.09
0.3%
220
+0.07
0.2%
1387
+0.07
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1752
+0.09
0.03
603
+0.07
0.03
628
+0.07
0.03
Negative Logits
<bos>
-1.58
-0.75
-0.73
-0.72
-0.72
-0.72
-0.71
-0.71
-0.71
-0.70
POSITIVE LOGITS
Juf
1.96
increa
1.84
fta
1.81
impra
1.78
affor
1.77
ftu
1.75
aen
1.73
fup
1.73
maneu
1.73
accla
1.72
Activations Density 0.122%