INDEX
Explanations
words related to economics and trade
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1343
+0.25
0.8%
1577
+0.17
0.5%
1842
+0.14
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1577
+0.25
0.07
1510
+0.17
0.05
1842
+0.14
0.05
Negative Logits
reluct
-1.67
shenan
-1.53
depic
-1.50
philanth
-1.48
Confu
-1.47
encomp
-1.47
snoopy
-1.43
disagre
-1.43
milf
-1.42
strick
-1.40
POSITIVE LOGITS
etc
0.98
etc
0.89
usw
0.79
等等
0.75
itp
0.75
등
0.59
cetera
0.58
OSSARY
0.56
等
0.54
ועוד
0.54
Activations Density 0.412%