INDEX
Explanations
terms related to spirals or spiral patterns
New Auto-Interp
Negative Logits
oose
-0.17
ymes
-0.16
ouns
-0.15
Karel
-0.15
ustry
-0.15
iterr
-0.15
ieties
-0.15
Kub
-0.15
coming
-0.15
693
-0.14
POSITIVE LOGITS
acular
0.15
urge
0.15
rega
0.15
iç
0.15
Shapiro
0.15
egal
0.14
eyer
0.14
Camden
0.14
acus
0.14
é«
0.14
Activations Density 0.013%