https://www.michaelaerni.com/publication-type/1/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/memorization/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/publication/modal-aphasia/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/multimodal-language-models/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/publication_types/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/safety/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tags/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/unified-representations/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/data-extraction/2025-03-03T23:48:08+01:00https://www.michaelaerni.com/tag/large-language-models/2025-03-03T23:48:08+01:00https://www.michaelaerni.com/publication/measuring-non-adversarial-reproduction-of-training-data-in-large-language-models/2025-03-03T23:48:08+01:00https://www.michaelaerni.com/tag/originality/2025-03-03T23:48:08+01:00https://www.michaelaerni.com/tag/privacy/2025-03-03T23:48:08+01:00https://www.michaelaerni.com/tag/auditing/2024-10-08T10:56:25+02:00https://www.michaelaerni.com/publication/evaluations-of-machine-learning-privacy-defenses-are-misleading/2024-10-08T10:56:25+02:00https://www.michaelaerni.com/tag/membership-inference/2024-10-08T10:56:25+02:00https://www.michaelaerni.com/tag/ml-privacy/2024-10-08T10:56:25+02:00https://www.michaelaerni.com/tag/benign-overfitting/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/deep-learning-theory/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/generalization-bounds/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/high-dimensional-statistics/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/tag/non-parametric-regression/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/publication/strong-inductive-biases-provably-prevent-harmless-interpolation/2026-01-28T07:58:02+01:00https://www.michaelaerni.com/publication/interpolation-can-hurt-robust-generalization-even-when-there-is-no-noise/2024-04-29T16:23:12+02:00https://www.michaelaerni.com/tag/learning-theory/2024-04-29T16:23:12+02:00https://www.michaelaerni.com/tag/regularization/2024-04-29T16:23:12+02:00https://www.michaelaerni.com/tag/robustness/2024-04-29T16:23:12+02:00https://www.michaelaerni.com/publication-type/8/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/tag/adversarial-robustness/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/tag/interpolation/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/tag/ridge-regularization/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/publication/surprising-benefits-of-ridge-regularization-for-noiseless-regression/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/tag/theory/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/tag/adversarial-training/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/publication/maximizing-the-robust-margin-provably-overfits-on-noiseless-data/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/tag/robust-overfitting/2021-08-15T22:59:34+02:00https://www.michaelaerni.com/categories/https://www.michaelaerni.com/publication/2021-01-23T19:43:06+00:00