INDEX
Explanations
references to news events or particular locations
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
856
+0.15
0.5%
906
+0.12
0.4%
1403
+0.10
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1415
+0.15
0.04
1134
+0.12
0.05
1843
+0.10
0.04
Negative Logits
setColor
-0.53
ennemi
-0.51
testSet
-0.49
createSlice
-0.49
isFirst
-0.47
释放
-0.46
işaret
-0.44
isSuccess
-0.44
başladı
-0.44
isActive
-0.43
POSITIVE LOGITS
abnorm
1.13
handels
1.11
peculi
1.10
glan
1.10
coö
1.08
dispen
1.08
logis
1.07
inder
1.06
socie
1.06
stoff
1.05
Activations Density 0.433%