INDEX
Explanations
various forms of the verb "to be" and auxiliary verbs related to actions and states
New Auto-Interp
Negative Logits
itar
-0.15
habi
-0.14
shouldn
-0.14
ivel
-0.14
HeaderComponent
-0.14
efault
-0.14
ht
-0.13
ARGER
-0.13
ULD
-0.13
ewan
-0.13
POSITIVE LOGITS
Here
0.21
Yet
0.21
away
0.20
However
0.20
here
0.19
Now
0.19
apart
0.19
yet
0.19
forth
0.19
n
0.19
Activations Density 0.018%