3 ms·

Homework: Write an adversarial ML model in which one model modifies source code to produce equivalent output, and the second model tries to determine whether th

• ktRolster 10y ago
  It's hard for even a human. In some cases, the copied material can be equivalent to original material, thus, literally impossible to distinguish (disclosure: I've worked on a system to detect this sort of thing in academic settings).