Metraipolis › archive › dc_fe4963c033a0

Kähler landscapes for complex neural network descents and guarantees including a search and destroy of the Calabi-Yau manifold — arXiv:2608.19584

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.19584
License (as archived)
CC0-1.0
Archived
2026-10-18T11:48:13.676Z
SHA-256 of the archived text
e906ff0ebc61f175298495ebe1440066877a32021f5708ad23e6de958a7016e1
Stable URL of this page
https://metraipolis.com/documento/dc_fe4963c033a0

How to cite

Kähler landscapes for complex neural network descents and guarantees including a search and destroy of the Calabi-Yau manifold — arXiv:2608.19584 — arXiv (en), https://arxiv.org/abs/2608.19584 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fe4963c033a0 (Metraipolis archive dc_fe4963c033a0, sha256 e906ff0ebc61f175298495ebe1440066877a32021f5708ad23e6de958a7016e1)
BibTeX
@misc{dcfe4963c033a0,
  title = {Kähler landscapes for complex neural network descents and guarantees including a search and destroy of the Calabi-Yau manifold — arXiv:2608.19584},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.19584},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fe4963c033a0; sha256 e906ff0ebc61f175298495ebe1440066877a32021f5708ad23e6de958a7016e1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.19584. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We study landscapes for complex-parameterized networks. Our approach is motivated with an information-theoretic manifold perspective of the parameter and via classical optimization guarantees although of complex geometric variety such as through Dolbeault asymptotics. The descent path admits a Kähler information metric under a cross-entropy via the Wirtinger Hessian on the log-likelihood potential. We restrict attention to a descent update rule with natural gradient descent via a differentiated loss scaled by the inverse metric, so the descent path remains in the holomorphic tangent bundle. We emphasize Calabi-Yau information manifolds which profane theoretical guarantees via an ill-curvature-conditioned landscape. We focus on Calabi-Yau metrics specifically in a non-compact setting with a global potential, so defined geometrically rather than invoking the topological requirements of the Calabi conjecture. In non-compact settings, we can write the metric determinant with respect to a background in terms of a pluriharmonic or real-valued function. Under bounded, nonuniform, and almost low-rank assumptions, we get a partial eigenvalue blow-up effect. In an empirical setting, a Ricci-flat metric will not form, but the blow-up effect is a local condition and can partially hold empirically on open sets. We isolate the Calabi-Yau case in a theoretical setting, and we counteract the corrupted geometries under regularization. Moreover, it has been discovered that negative curvature subverts the loss landscape, specifically sectional curvature, so we expand on this and draw interconnections to negative-definite Ricci curvature. Our arguments primarily exist via geometric analysis, although we establish roots in deep learning theory such as through asymptotics at initialization and connections through failure modes of neural network guarantees under vanishing and negative Ricci curvature.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-21 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Differential Geometry (math.DG); Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2608.19584 — 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 2026-10-18T11:47:43.305Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).