INDEX
Explanations
expressions of legal rights and obligations
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
486
+0.14
0.8%
423
+0.14
0.8%
407
+0.13
0.8%
Correlated Neurons
Index
P. Corr.
Cos Sim.
407
+0.14
0.18
423
+0.14
0.18
88
+0.13
0.12
Negative Logits
uscular
-1.70
ker
-1.53
loads
-1.51
³
-1.47
ride
-1.40
ften
-1.37
teral
-1.36
Scient
-1.36
no
-1.35
ly
-1.34
POSITIVE LOGITS
"}](#
1.85
coff
1.60
agna
1.57
googleapis
1.56
adoc
1.55
askell
1.49
ejemplo
1.42
'];
1.41
Papers
1.38
]'
1.35
Activations Density 3.278%