INDEX
Explanations
increment and decrement operators in code
New Auto-Interp
Negative Logits
rest
-0.60
„
-0.58
da
-0.56
pe
-0.55
St
-0.54
and
-0.53
del
-0.53
nem
-0.51
-0.51
Hold
-0.50
POSITIVE LOGITS
++
3.28
++
2.58
++,
2.46
++
2.36
++.
2.30
++$
2.17
++]
2.05
(++
2.04
++)
1.97
(++
1.94
Activations Density 0.094%