Binary, octal, decimal, and hexadecimal are ways to write the same integer. Decimal 26 is binary 11010, octal 32, and hex 1A. Choose the source and destination bases explicitly, then test a known value and convert it back.
Check how the tool handles negative numbers, leading zeros, large integers, and fractional values. Some implementations are limited by floating-point precision, so long identifiers may be rounded. Prefixes such as 0x are notation conventions, not part of the numeric value. For byte-level work, specify fixed width and signedness separately from the base.