INDEX
Explanations
food-related items or recipes, such as specific ingredients and cooking instructions
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.18
1.0%
2034
+0.12
0.7%
1334
+0.10
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1385
+0.18
0.15
2034
+0.12
0.13
1870
+0.10
0.06
Negative Logits
<bos>
-2.91
ⓧ
-1.14
/***
-0.97
/**
-0.97
-0.94
Примеча
-0.89
Vegeu
-0.86
<?
-0.77
webElementXpaths
-0.76
solidar
-0.74
POSITIVE LOGITS
paradiso
1.27
signora
1.17
tramonto
1.14
bandung
1.12
jaya
1.08
riva
1.08
bosco
1.08
pioggia
1.08
maroc
1.07
cavallo
1.07
Activations Density 1.926%