> ML_LITERATURE // GOODFELLOW-2014-EXPLAINING-AND-HARNESSING-ADVERSARIAL-EXAMPLES_v1.0
Explaining and Harnessing Adversarial Examples (FGSM)
Ian J. Goodfellow, Jonathon Shlens, Christian Szegedy · International Conference on Learning Representations (ICLR) (2014)
safety-fairness2014industry-standardartifactsAvailable
Principal Contribution
Explained adversarial vulnerability as a consequence of high-dimensional linearity; introduced Fast Gradient Sign Method (FGSM).
Operational Relevance
Directly guides deployment choices and architecture selection for task-adversarial-robustness.
Assumptions
- Standard empirical regularity and statistical stability hold across evaluation domains
Limitations
- Performance characteristics depend on domain distribution and compute allocation parameters
Connected Algorithms, Architectures & Tools
Related Algorithms:
Related Architectures:
Implementing Libraries: