INDEX
Explanations
references to specific company names or product names
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
283
+0.13
0.4%
1741
+0.13
0.4%
304
+0.11
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1984
+0.13
0.06
227
+0.13
0.07
283
+0.11
0.01
Negative Logits
shewn
-0.84
Datuak
-0.82
különböz
-0.73
blest
-0.70
WriteBarrier
-0.65
SPECTION
-0.63
hozz
-0.63
friable
-0.62
besz
-0.62
szüks
-0.61
POSITIVE LOGITS
gesta
1.25
Nö
1.24
frau
1.23
plak
1.23
utop
1.20
glan
1.20
drap
1.20
elek
1.18
logis
1.17
biograf
1.16
Activations Density 0.339%