INDEX
Explanations
keynote speaker information from event summaries
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1708
+0.10
0.3%
605
+0.08
0.2%
1262
+0.07
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1363
+0.10
0.04
690
+0.08
0.05
783
+0.07
0.04
Negative Logits
Enum
-0.57
Ty
-0.56
numel
-0.55
ima
-0.54
Ty
-0.54
Lock
-0.53
obj
-0.53
O
-0.53
obj
-0.50
Loja
-0.50
POSITIVE LOGITS
keynote
1.29
Keynote
1.22
beverly
1.20
veneta
1.18
fte
1.18
jaya
1.17
toscana
1.17
signora
1.14
😭😭
1.13
dises
1.11
Activations Density 0.322%