This document discusses extended differential fuzzing techniques. It begins with an overview of common fuzzing and differential fuzzing. Extended differential fuzzing aims to detect more vulnerability types by analyzing outputs across different implementations, inputs, versions, and operating systems. These include path disclosure, user disclosure, error disclosure, code evaluation, command execution, network connections, and file reads. The document demonstrates examples of detecting these behaviors in PHP, Perl, Python and other languages. It promotes an open source fuzzing framework called XDiFF that automates extended differential analysis.