Results for “malcolm-x”

Detect poisoned training data and backdoored models across the ML pipeline using statistical analysis, activation clustering, and spectral signatures.

Extracts and analyzes malicious VBA macros, XLM macros, DDE, and remote template injections in Microsoft Office documents using olevba, oledump, and deobfuscation techniques to identify download cradles, payload execution, and persistence mechanisms.