aerOS project retweetledi

🔐 New from #ANTIDOTEproject: Our latest study tests how adversarial attacks (FGSM, PGD, DeepFool, C&W) affect popular AI models — revealing that bigger models aren’t always more robust.
🔗 zenodo.org/records/177080…
#AdversarialAI #AISecurity #TrustworthyAI
@6G_SNS @HorizonEU

English




























