Closed mhmd97z closed 3 months ago
To save the counterexample, set the following options: https://github.com/Verified-Intelligence/alpha-beta-CROWN/blob/main/complete_verifier/arguments.py#L87-L89 https://github.com/Verified-Intelligence/alpha-beta-CROWN/blob/main/complete_verifier/arguments.py#L772-L774
Hello, how to generate the exact adversarial example when the output's upper bound is negative in input or relu splitting?