TRIPLE DESCENT AND THE TWO KINDS OF OVERFITTING: WHERE & WHY DO THEY APPEAR?

von D'Ascoli, Stéphane
Veröffentlicht in NeurIPS (34. : 2020 : Online) 34th Conference on Neural Information Processing Systems (NeurIPS 2020) ; Volume 4 of 27 (2021)

ALIGN, THEN MEMORISE: THE DYNAMICS OF LEARNING WITH FEEDBACK ALIGNMENT

von REFINETTI, MARIA
Veröffentlicht in International Conference on Machine Learning (38. : 2021 : Online) International Conference on Machine Learning (ICML 2021 ; Part 12 of 16 (2022)

DOUBLE TROUBLE IN DOUBLE DESCENT: BIAS AND VARIANCE(S) IN THE LAZY REGIME

von d'Ascoli, Stéphane
Veröffentlicht in International Conference on Machine Learning (37. : 2020 : Online) 37th International Conference on Machine Learning (ICML 2020) ; Part 3 of 15 (2021)

Conditioned Text Generation with Transfer for Closed-Domain Dialogue Systems

von d’Ascoli, Stéphane
Veröffentlicht in SLSP (8. : 2020 : Cardiff) Statistical language and speech processing (2020)

CONVIT: IMPROVING VISION TRANSFORMERS WITH SOFT CONVOLUTIONAL INDUCTIVE BIASES

von D’ASCOLI, STÉPHANE
Veröffentlicht in International Conference on Machine Learning (38. : 2021 : Online) International Conference on Machine Learning (ICML 2021 ; Part 3 of 16 (2022)

On the interplay between data structure and loss function in classification problems

von d' Ascoli, Stéphane
Veröffentlicht in NeurIPS (35. : 2021 : Online) 35th Conference on Neural Information Processing Systems (NeurIPS 2021) ; Volume 11 of 36 (2022)

End-to-end Symbolic Regression with Transformers

von Kamienny, Pierre-Alexandre
Veröffentlicht in NeurIPS (36. : 2022 : New Orleans, La.; Online) 36th Conference on Neural Information Processing Systems (NeurIPS 2022 ; Volume 14 of 50 (2023)