Metraipolis › archive › dc_4ef6ea4ac772
Momentum as Residual-Driven Multiplier Correction for Deep Learning Optimization — arXiv:2608.12925
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.12925
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T04:27:29.420Z
- SHA-256 of the archived text
6185c03bd9a30021dfcbcec244a56b0faa64b0a1703036bcee129e347d7de0cd- Stable URL of this page
https://metraipolis.com/documento/dc_4ef6ea4ac772
How to cite
Momentum as Residual-Driven Multiplier Correction for Deep Learning Optimization — arXiv:2608.12925 — arXiv (en), https://arxiv.org/abs/2608.12925 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4ef6ea4ac772 (Metraipolis archive dc_4ef6ea4ac772, sha256 6185c03bd9a30021dfcbcec244a56b0faa64b0a1703036bcee129e347d7de0cd)
BibTeX
@misc{dc4ef6ea4ac772,
title = {Momentum as Residual-Driven Multiplier Correction for Deep Learning Optimization — arXiv:2608.12925},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.12925},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4ef6ea4ac772; sha256 6185c03bd9a30021dfcbcec244a56b0faa64b0a1703036bcee129e347d7de0cd},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.12925. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Momentum-based optimizers are widely used in modern deep learning, yet the relations among momentum recursion, update geometry, and acceleration remain only partially understood. We develop an $\textbf{A}$DMM-$\textbf{I}$nspired $\textbf{M}$omentum (AIM) framework based on residual-penalty variable splitting, which interprets momentum as a multiplier-like correction driven by the splitting residual. AIM recovers the exponential moving average of gradients from an ADMM-style multiplier update and separates two mechanisms that are usually intertwined in practical optimizers: the residual penalty determines the update geometry, whereas the approximation of the objective-related subproblem determines the acceleration form. Building on AIM, we propose $\textbf{R}$elativistic $\textbf{A}$daptive gradient $\textbf{D}$escent with $\textbf{A}$ccelerated $\textbf{R}$esidual (RADAR), which combines relativistic adaptive geometry, decoupled residual correction, and second-order momentum filtering to improve the update direction and momentum estimation. We establish stochastic convergence through a variance-perturbed Lyapunov drift analysis. Experiments on supervised vision learning, language modeling, and reinforcement learning show that RADAR achieves consistent improvements over strong adaptive optimizer baselines.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.12925 — 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 2026-10-19T04:20:35.834Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).