model security
AI security interview questions tagged model security, across every topic.
3 questions · 2 unlocked for you
Concepts behind "model security"
Explanations of the mechanisms these questions rely on.
Foundational
Unsafe Model Files: Pickle versus SafetensorsMany machine-learning checkpoints have been saved with Python's pickle, a format whose loader can execute code chosen by the file's author, so loading an untrusted model can compromise the machine. Safetensors stores only tensors and a small JSON header, so loading cannot run code. It fixes that one problem: it does not detect backdoored weights, malicious custom code shipped alongside a model, or poisoned training. A safe model intake process combines safe formats, restricted loaders, scanning, provenance and isolation. Supply Chain Security
Core
Backdoors and Trojaned ModelsA backdoored model behaves normally except when a specific trigger appears in its input, at which point it does what the attacker chose. Backdoors are implanted through poisoned training or fine-tuning data or by directly modifying weights, and they reach victims through shared pre-trained models and datasets. They are hard to find because testing rarely includes the trigger, and research has shown some backdoors survive later safety training. Provenance, controlled fine-tuning, evaluation and architectural containment matter more than detection. Adversarial ML & PrivacySign in