INDEX
Explanations
structured data representations and procedural patterns in text
New Auto-Interp
Negative Logits
ifo
-0.14
ones
-0.14
ONES
-0.14
obec
-0.14
ĵ¨
-0.13
ahn
-0.13
so
-0.13
andy
-0.13
наÑĥк
-0.13
abs
-0.13
POSITIVE LOGITS
jedn
0.17
nodoc
0.15
AllWindows
0.15
fwd
0.15
#endregion
0.14
DropIndex
0.14
#End
0.14
Erotische
0.14
TOTAL
0.14
//</
0.14
Activations Density 0.212%