AI Paper Atlas

Research map / Optimization / Nonconvex and smooth optimization

A Stability Analysis of AdamW: Unstable Equilibria and Non-Convergence

Baiyu Su, Lizhang Chen, Qiang Liu

NeurIPS 2026 · Optimization › Nonconvex and smooth optimization · approach: Coordinate and Adam convergence analysis · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Theoretical dynamical-systems analysis of AdamW showing its fixed points correspond to stationary points of a constrained, regularized problem, with only some being stable and non-convergence possible.

Problem

Understanding the behavior of AdamW, which is widely used in deep learning but poorly understood theoretically, particularly its convergence and the stability of its fixed points.

Approach

Treats AdamW as a dynamical system and shows it admits an implicit fixed-point objective whose fixed points are stationary points of a constrained and regularized optimization problem; stability of these fixed points depends on hyperparameters, and some fixed points are unstable, implying non-convergence.

Key results

Measured effects reported in the paper

What was changedWhereMeasuredEffect
rescale AdamW weight decay by (|m|+e)/(sqrt(v)+e) (AdamMe)AdamW · GPT-2-Small (124M)validation loss: 3.355 -> 3.317Improved
use beta1 > beta2 with quadratic objective f(x)=(x-0.5)^2AdamW · 1D toy objectivestability of equilibrium: requires f''<9.1e-2 but f''=2Hurt
use beta1=0.9, beta2=0.99 with non-convex quarticAdamW · 1D toy objectiveconvergence: f''(x*)=-1.20 vs required > -0.22Hurt
set beta1 = beta2 (equal momentum coefficients) (ablation)AdamW · 1D stability analysisstability conditionImproved
rescale weight decay (AdamMe) with b<=4aAdamW ODE · continuous-time dynamicsconvergence guaranteeImproved

Surprising findings

Limitations

Experimental setup

Builds on: AdamW

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy