INDEX
Explanations
ideas related to centering, focus, and themes of organization in discussions
New Auto-Interp
Negative Logits
̧
-0.14
jer
-0.14
ais
-0.14
idenav
-0.14
idia
-0.14
inqu
-0.14
/*!
-0.14
Sab
-0.13
sidel
-0.13
tero
-0.13
POSITIVE LOGITS
upon
0.24
less
0.24
around
0.23
heavily
0.22
around
0.20
-around
0.19
upon
0.19
Upon
0.18
entirely
0.18
partly
0.17
Activations Density 0.109%