INDEX
Explanations
instances of time-related phrases and mentions of specific names
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1081
+0.07
0.2%
856
+0.07
0.2%
1226
+0.07
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
693
+0.07
0.03
1081
+0.07
0.04
59
+0.07
0.03
Negative Logits
encomp
-1.08
impra
-1.05
shenan
-1.04
indescri
-1.04
increa
-1.03
unspeak
-1.01
disagre
-0.99
intersper
-0.98
affor
-0.95
reluct
-0.94
POSITIVE LOGITS
Spisak
0.59
LOAT
0.59
AssemblyProduct
0.57
PARSER
0.56
RTSC
0.55
はじめに
0.55
विश्वसनीयता
0.54
ValueStyle
0.54
للاسماء
0.53
sphase
0.53
Activations Density 0.279%