Metraipolis › archive › dc_8c4dca32bfcf
Curvature-Aware Zeroth-Order Optimization for Memory-Efficient Test-Time Adaptation — arXiv:2608.12279
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.12279
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T05:49:44.955Z
- SHA-256 of the archived text
f0ac4dfe550c15b73a9c0944fce821da58700ed37bcc42b7694d0072d08411ad- Stable URL of this page
https://metraipolis.com/documento/dc_8c4dca32bfcf
How to cite
Curvature-Aware Zeroth-Order Optimization for Memory-Efficient Test-Time Adaptation — arXiv:2608.12279 — arXiv (en), https://arxiv.org/abs/2608.12279 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8c4dca32bfcf (Metraipolis archive dc_8c4dca32bfcf, sha256 f0ac4dfe550c15b73a9c0944fce821da58700ed37bcc42b7694d0072d08411ad)
BibTeX
@misc{dc8c4dca32bfcf,
title = {Curvature-Aware Zeroth-Order Optimization for Memory-Efficient Test-Time Adaptation — arXiv:2608.12279},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.12279},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8c4dca32bfcf; sha256 f0ac4dfe550c15b73a9c0944fce821da58700ed37bcc42b7694d0072d08411ad},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.12279. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Test-time adaptation (TTA) aims to enhance the cross-domain performance of pre-trained models by adapting to unlabeled test data. While most existing TTA methods rely on backpropagation (BP) for finetuning, BP-free methods such as zeroth-order (ZO) methods are more desired in practical on-device scenarios. ZO methods rely only on forward computation, which can largely reduce the complexity and memory overhead of on-device deployment. However, ZO methods suffer from much higher variance compared with first-order methods in estimating the gradient. To address this, we propose an improved ZO method to substantially boost the performance of ZO optimization based TTA. First, we provide an observation to reveal the persistent low-rank Hessian structure of the loss during the adaptation process. Based on this insight, we then propose a loss-landscape curvature-aware zeroth-order (CAZO) method, which leverages a sliding-average estimation of the diagonal Hessian to construct a covariance matrix for anisotropic perturbation sampling. CAZO operates by freezing pretrained weights and optimizing minimal adapter parameters via forward-only passes based gradient estimation, which can substantially reduce the memory overhead compared to BP-based methods. Extensive experiments demonstrate that CAZO significantly outperforms existing TTA methods, achieving state-of-the-art performance while maintaining an excellent balance between accuracy and memory efficiency. Code is available at this https URL . Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-13 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2608.12279 — 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 2026-10-19T05:42:25.927Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-13?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).