Linked Presentation: Aliasing Backdoor Attacks on Pre-trained ModelsDifferential Testing of Cross Deep Learning Framework APIs: Revealing Inconsistencies and Vulnerabilities
