12 articles in this category
Adversarial training is the default answer whenever someone asks how to harden a model against attacks. You add perturbed examples to the training set...
Imagine you're shipping a model that must survive both white-box and black-box attacks. White-box attackers walk through your gradients like an unlock...
You train a model. Run PGD-40. Accuracy holds at 85%. Feels good. Then someone tries an adaptive attack — maybe a different loss, a random restarts, a...
So you've tested your model against FGSM, PGD, maybe even AutoAttack. Low robust accuracy. Nice. But here's the dirty secret: standard evaluations mis...
You've built a classifier that hits 94% on clean validation. Then someone whispers: what if an attacker tweaks a pixel? So you add adversarial trainin...
Semantic adversarial examples exploit the gap between mathematical guarantees and human perception. A model certified robust against all ℓ∞...
You trained your model. It hits 98% test accuracy. You deploy it. Then someone puts a sticker on a stop sign, and your car thinks it's a speed limit. ...
You trained a model. It hits 98% probe accuracy. PGD with epsilon=8/255 barely drops it to 92%. You think: robust enough for deployment. Not always tr...
Certified robustness guarantee sound impressive on paper. They promise that for any L p -bounded perturbaal, your model's prediction stays put. But he...
You spent weeks hardening a model. Added adversarial trained, defensive distillation, maybe even a certified defense. Then someone runs an adaptive at...
If you have ever uploaded a student project to a robustness leaderboard, you have probably trained against PGD-ℓ∞ with epsilon 8/255. It is the defaul...
Adversarial robustness and model calibration rarely share a headline. Most practitioners chase clean accuracy under attack, ignoring whether the model...