INDEX
Explanations
phrases related to baking and cooking
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1527
+0.14
0.6%
120
+0.13
0.6%
1637
+0.13
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1520
+0.14
0.02
201
+0.13
0.01
120
+0.13
0.02
Negative Logits
Quinoa
-0.62
Pesto
-0.59
Smoothie
-0.55
nutella
-0.54
Zucchini
-0.53
ഔ
-0.52
Cauliflower
-0.49
ఊ
-0.49
Cabbage
-0.47
bonbons
-0.46
POSITIVE LOGITS
Baker
1.27
bake
1.25
baker
1.21
Baker
1.18
baked
1.15
baking
1.15
bakers
1.13
bake
1.13
BAK
1.11
baker
1.07
Activations Density 0.103%