Metraipolis › archive › dc_cc4ef6912803

Exploring Bad Minima in Transfer Learning — arXiv:2610.41807

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.41807
License (as archived)
CC0-1.0
Archived
2026-11-03T08:10:52.765Z
SHA-256 of the archived text
ab96dfe30be6f8ff52026ca7434af08a06184c82fc03304d4115bc9f39774309
Stable URL of this page
https://metraipolis.com/documento/dc_cc4ef6912803

How to cite

Exploring Bad Minima in Transfer Learning — arXiv:2610.41807 — arXiv (en), https://arxiv.org/abs/2610.41807 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cc4ef6912803 (Metraipolis archive dc_cc4ef6912803, sha256 ab96dfe30be6f8ff52026ca7434af08a06184c82fc03304d4115bc9f39774309)
BibTeX
@misc{dccc4ef6912803,
  title = {Exploring Bad Minima in Transfer Learning — arXiv:2610.41807},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.41807},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cc4ef6912803; sha256 ab96dfe30be6f8ff52026ca7434af08a06184c82fc03304d4115bc9f39774309},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.41807. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Transfer learning is a highly successful learning paradigm in deep learning which leverages knowledge from one task and applies it to another. Despite its success, the necessary conditions for knowledge transfer are unclear and require further research. In this dissertation, our aim is to understand transferability a little better by answering the following question: Does good generalization on a source task necessarily imply transferability to a related target task? We hypothesize the answer is negative, and attempt to find counterexamples via adversarial attacks on the transfer learning procedure. Over three types of adversarial attacks, our empirical results show a strong relationship between generalization on the source task and transferability. These results do not rule out generalization on the source task as a sufficient condition for successful transfer.

Tema degli esploratori: fiducia digitale e crittografia (elenco arXiv cs.CR, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Cryptography and Security (cs.CR).
Pagina del lavoro: https://arxiv.org/abs/2610.41807 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-03T08:02:08.788Z dalla pagina https://arxiv.org/list/cs.CR/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).