INDEX
Explanations
numeric values or words related to numbers or quantities
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
776
+0.19
0.6%
1177
+0.15
0.5%
381
+0.13
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
776
+0.19
0.06
1731
+0.15
0.02
194
+0.13
0.02
Negative Logits
بيها
-0.65
createContext
-0.64
gyerek
-0.59
tramonto
-0.57
merie
-0.55
mattino
-0.55
枚目
-0.55
useCallback
-0.53
íteni
-0.53
jectures
-0.53
POSITIVE LOGITS
Juf
1.05
Bartholo
1.02
Rine
1.02
Gorb
1.00
depic
0.96
Kün
0.96
viendra
0.91
accla
0.91
préc
0.91
McLaugh
0.91
Activations Density 0.169%