INDEX
Explanations
HTML heading tags and their hierarchical structure
New Auto-Interp
Negative Logits
↵↵
-0.66
.
-0.61
<eos>
-0.57
(
-0.56
-0.52
dex
-0.50
ani
-0.50
↵
-0.49
;
-0.48
o
-0.48
POSITIVE LOGITS
iſt
1.14
ainfi
1.11
ſy
0.98
auffi
0.98
itſelf
0.98
plufieurs
0.97
Efq
0.96
auroit
0.96
nahilalakip
0.95
Reſ
0.94
Activations Density 0.039%