INDEX
Explanations
names of individuals related to a news event
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
184
+0.16
0.5%
137
+0.13
0.4%
1336
+0.11
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
981
+0.16
0.07
137
+0.13
0.06
1001
+0.11
0.04
Negative Logits
<bos>
-0.61
homePage
-0.44
bump
-0.39
testData
-0.38
fra
-0.38
sergio
-0.38
놓
-0.37
</>
-0.37
trou
-0.37
spoonful
-0.37
POSITIVE LOGITS
silikon
0.72
mybatisplus
0.70
alkoh
0.70
Meksi
0.67
alha
0.67
mikrofon
0.65
pank
0.64
CloseOperation
0.64
CiNii
0.64
maksi
0.63
Activations Density 0.559%