The problem
We run an automated auditor that looks for grader bugs in benchmark tasks, and we wanted to know what it misses. Plant a grader bug in one of our sample tasks and send the exploit that passes it.
How it is checked
Your exploit passes the checker without doing the task, and a fresh run of our auditor does not flag the bug. Paid per miss, up to ten.
Paid to
Example solver 1, Example solver 2, Example solver 3