INDEX
Explanations
financial terms and numerical values
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
304
+0.13
0.5%
678
+0.13
0.4%
1842
+0.12
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
755
+0.13
0.03
281
+0.13
0.03
501
+0.12
0.03
Negative Logits
on
-0.85
in
-0.85
to
-0.84
for
-0.84
so
-0.83
of
-0.82
but
-0.82
at
-0.82
just
-0.81
that
-0.81
POSITIVE LOGITS
fta
2.19
ftu
2.14
embra
2.08
vns
2.07
accla
2.04
nece
2.02
fep
1.99
desir
1.97
paff
1.95
mef
1.94
Activations Density 0.076%