TAROT derives a new robust margin disparity generalization bound and uses it to train domain-invariant, adversarially robust classifiers that outperform prior robust UDA methods on standard benchmarks.
Spectrally-normalized margin bounds for neural networks
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
TAROT: Towards Essentially Domain-Invariant Robustness with Theoretical Justification
TAROT derives a new robust margin disparity generalization bound and uses it to train domain-invariant, adversarially robust classifiers that outperform prior robust UDA methods on standard benchmarks.