INDEX
Explanations
phrases describing tearing or ripping something apart
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
889
+0.15
0.5%
866
+0.15
0.5%
251
+0.14
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
866
+0.15
0.02
889
+0.15
0.02
251
+0.14
0.02
Negative Logits
ModelAdmin
-0.48
Catégorie
-0.47
FontStyle
-0.47
ChildScrollView
-0.45
elemField
-0.44
Galería
-0.43
InputTagHelper
-0.43
FlatStyle
-0.43
Wicidata
-0.43
cedido
-0.42
POSITIVE LOGITS
tearing
1.14
tear
1.11
tore
1.07
torn
1.02
Torn
1.01
Tear
0.99
rips
0.94
ripped
0.94
Tear
0.93
rip
0.93
Activations Density 0.076%