Metraipolis › archive › dc_b35fae0bbf20

Quadratic Weak-to-Strong Generalization in Random Feature Networks via Random Matrix Theory — arXiv:2610.09044

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.09044
License (as archived)
CC0-1.0
Archived
2026-10-08T12:36:40.967Z
SHA-256 of the archived text
149ce8e35b894e1f552014bd520cc51c79b222472c248e453db85e6b74ab8905
Stable URL of this page
https://metraipolis.com/documento/dc_b35fae0bbf20

How to cite

Quadratic Weak-to-Strong Generalization in Random Feature Networks via Random Matrix Theory — arXiv:2610.09044 — arXiv (en), https://arxiv.org/abs/2610.09044 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b35fae0bbf20 (Metraipolis archive dc_b35fae0bbf20, sha256 149ce8e35b894e1f552014bd520cc51c79b222472c248e453db85e6b74ab8905)
BibTeX
@misc{dcb35fae0bbf20,
  title = {Quadratic Weak-to-Strong Generalization in Random Feature Networks via Random Matrix Theory — arXiv:2610.09044},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.09044},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b35fae0bbf20; sha256 149ce8e35b894e1f552014bd520cc51c79b222472c248e453db85e6b74ab8905},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09044. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Weak-to-strong generalization is the phenomenon where a strong student model trained with labels produced by a weak teacher model is able to generalize better than the teacher. In this paper, we study this phenomenon in two-layer random feature networks where the model strength is determined by its width. Using tools from random matrix theory, we derive deterministic equivalents for the population errors of an optimally trained teacher and a student trained with gradient flow. For ReLU activation and a pure spherical harmonic target, we obtain sharp asymptotics under a Gaussian universality assumption, showing a quadratic improvement: the student error scales as the square of the teacher error. These results attain the general lower bound of Medvedev at al (2025). We also analyze how the student behaves under more general stopping times and targets supported on multiple harmonic degrees, characterizing the regimes in which weak-to-strong generalization occurs and identifying the transition between quadratic, non-quadratic, and no improvement.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG); Probability (math.PR).
Pagina del lavoro: https://arxiv.org/abs/2610.09044 — 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 2026-10-08T12:14:15.371Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).