INDEX
Explanations
phrases where something positive is followed by a connector indicating further positive actions or consequences
the conjunction "And" in various contexts
New Auto-Interp
Negative Logits
normally
-0.66
pson
-0.62
due
-0.60
rue
-0.58
prior
-0.57
alike
-0.57
approx
-0.56
preferred
-0.56
useful
-0.55
eligibility
-0.55
POSITIVE LOGITS
And
2.83
And
2.14
But
1.63
Which
1.51
So
1.50
Of
1.46
Or
1.46
Moreover
1.42
Then
1.42
That
1.41
Activations Density 0.052%