The syntax of the numeric format string is identical to the format string that the C printf function accepts, with the following exceptions:
Place a $ character after the % character in the format string to remove trailing zeros after the decimal point.
Use a format code of %b to format a number in binary.
Specify an empty string to set the format to the default numeric format.
I found the above in the help for PropertyObject.NumericFormat. Ideally, it would have also been in the help for the Numeric Format Dialog. I'll submit a request for that.