INDEX
Explanations
legal and ethical dilemmas, particularly related to reporting concerns in a workplace setting
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1967
+0.08
0.2%
1919
+0.08
0.2%
378
+0.07
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1919
+0.08
0.06
1415
+0.08
0.03
264
+0.07
0.03
Negative Logits
silikon
-1.14
antik
-1.09
keramik
-1.09
kafe
-1.06
optik
-1.06
kosme
-1.00
krim
-0.99
mikrofon
-0.99
karton
-0.98
teras
-0.95
POSITIVE LOGITS
cushi
0.81
naturally
0.78
unavoid
0.73
suscep
0.72
swee
0.66
plenti
0.65
ecru
0.65
disreg
0.64
snoopy
0.63
indescri
0.63
Activations Density 0.504%