INDEX
Explanations
words related to an audience
references to an audience
New Auto-Interp
Negative Logits
idy
-0.73
Plum
-0.65
Scot
-0.64
imov
-0.63
erald
-0.63
Baker
-0.62
Syn
-0.61
borg
-0.60
Franch
-0.60
phalt
-0.58
POSITIVE LOGITS
members
0.91
member
0.91
members
0.89
tuning
0.82
audience
0.81
participation
0.80
iences
0.79
tuned
0.79
Reviewer
0.79
ele
0.76
Activations Density 0.040%