INDEX
Explanations
Java programming constructs and related libraries
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
156
+0.24
1.4%
181
+0.13
0.7%
93
+0.12
0.7%
Correlated Neurons
Index
P. Corr.
Cos Sim.
181
+0.24
0.06
250
+0.13
0.05
80
+0.12
0.04
Negative Logits
its
-1.73
successfully
-1.60
reproduced
-1.53
reproduce
-1.53
]
-1.51
denly
-1.50
cludes
-1.48
↵
-1.45
ccess
-1.42
regards
-1.41
POSITIVE LOGITS
Ļª
3.29
ĨĴ
3.19
ĭ
3.14
↵
3.13
3.13
<|outofrange|>
3.13
č↵
3.13
3.13
3.13
<|outofrange|>
3.13
Activations Density 0.307%