A poisoning technique that keeps examples looking normal but assigns them the wrong label, shifting what the model learns to predict.
Continue to AI University →