AI Navigate

高度退化多項式に対するAdam収束性の理解に向けて

arXiv cs.LG / 2026/3/11

Ideas & Deep AnalysisModels & Research

要点

  • 本論文では、外部スケジューラや$\beta_2$が1に近いことに依存せず、特殊な高度に退化した多項式目的関数のクラスに対するAdam最適化器の自然な自己収束性を探求する。
  • 著者らは局所漸近安定性の理論的条件を導出し、Adamがこれらの退化多項式に対して局所線形収束を達成し、勾配降下法やモメンタムの亜線形収束速度を大幅に上回ることを示す。
  • Adamの収束加速は、2次モーメント推定値と勾配の2乗の間のデカップリングに起因し、これが実効学習率を指数的に増幅する。
  • さらに研究はAdamのハイパーパラメータ相図を特徴付け、安定収束、スパイク、SignGD類似の振動の3つの異なる振る舞い領域を特定し、Adamパラメータのより良い調整への洞察を提供する。
  • 実験結果は理論的発見と強く一致し、この文脈におけるAdamの挙動に関する提案された解析フレームワークを検証している。

Computer Science > Machine Learning

arXiv:2603.09581 (cs)
[Submitted on 10 Mar 2026]

Title:Towards Understanding Adam Convergence on Highly Degenerate Polynomials

View a PDF of the paper titled Towards Understanding Adam Convergence on Highly Degenerate Polynomials, by Zhiwei Bai and 4 other authors
View PDF HTML (experimental)
Abstract:Adam is a widely used optimization algorithm in deep learning, yet the specific class of objective functions where it exhibits inherent advantages remains underexplored. Unlike prior studies requiring external schedulers and $\beta_2$ near 1 for convergence, this work investigates the "natural" auto-convergence properties of Adam. We identify a class of highly degenerate polynomials where Adam converges automatically without additional schedulers. Specifically, we derive theoretical conditions for local asymptotic stability on degenerate polynomials and demonstrate strong alignment between theoretical bounds and experimental results. We prove that Adam achieves local linear convergence on these degenerate functions, significantly outperforming the sub-linear convergence of Gradient Descent and Momentum. This acceleration stems from a decoupling mechanism between the second moment $v_t$ and squared gradient $g_t^2$, which exponentially amplifies the effective learning rate. Finally, we characterize Adam's hyperparameter phase diagram, identifying three distinct behavioral regimes: stable convergence, spikes, and SignGD-like oscillation.
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2603.09581 [cs.LG]
  (or arXiv:2603.09581v1 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2603.09581
Focus to learn more
arXiv-issued DOI via DataCite

Submission history

From: Zhiwei Bai [view email]
[v1] Tue, 10 Mar 2026 12:30:20 UTC (6,902 KB)
Full-text links:

Access Paper:

Current browse context:
cs.LG
< prev   |   next >
Change to browse by:
cs

References & Citations

export BibTeX citation Loading...

BibTeX formatted citation

×
Data provided by:

Bookmark

BibSonomy logo Reddit logo
Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
Links to Code Toggle
Papers with Code (What is Papers with Code?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos

Demos

Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers

Recommenders and Search Tools

Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
IArxiv recommender toggle
IArxiv Recommender (What is IArxiv?)
About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.