INDEX
Explanations
verbs related to starting or undertaking tasks or projects
New Auto-Interp
Negative Logits
verbs
-0.70
gnu
-0.70
appa
-0.66
creen
-0.66
ggies
-0.59
karma
-0.58
acea
-0.57
cell
-0.57
acca
-0.56
stuff
-0.56
POSITIVE LOGITS
upon
0.92
TAIN
0.88
edIn
0.87
itect
0.80
ngth
0.79
antly
0.79
ments
0.75
eenth
0.75
ead
0.74
prise
0.74
Activations Density 0.029%