INDEX
Explanations
descriptions of construction work incidents
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1013
+0.17
0.5%
1120
+0.15
0.5%
939
+0.12
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
939
+0.17
0.06
1048
+0.15
0.02
1013
+0.12
0.06
Negative Logits
impra
-0.98
unspeak
-0.96
naï
-0.95
indescri
-0.90
intrigu
-0.90
intersper
-0.90
cuck
-0.90
shenan
-0.87
horrend
-0.87
attemp
-0.86
POSITIVE LOGITS
construction
0.76
contractors
0.70
contractor
0.69
Construction
0.66
施工
0.64
construction
0.63
Construction
0.59
crews
0.58
تكبرها
0.58
CONSTRUCTION
0.58
Activations Density 0.502%