INDEX
Explanations
the presence of the substring "em" across various contexts
New Auto-Interp
Negative Logits
ple
-0.51
adag
-0.49
ighted
-0.48
oinette
-0.48
helicopter
-0.47
AutoresizingMask
-0.47
Another
-0.46
findOne
-0.46
matter
-0.46
Gruber
-0.45
POSITIVE LOGITS
Em
1.06
Em
1.05
em
0.99
tagHelperRunner
0.98
Guild
0.89
em
0.89
@[+][
0.77
guild
0.72
enabled
0.72
EM
0.72
Activations Density 0.090%