AI Paper Atlas

Research map / Optimization / Optimizers

EIPO: Efficient Inter-Step Parallel Optimization

Jianrong Lu, Zhuoya Gu, Zhiyu Zhu, Hui LIU, Junhui Hou

NeurIPS 2026 · Optimization › Optimizers · approach: Adam convergence and adaptive learning rates

Read the paperCheck your idea against 137,217 papers

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

Recasts gradient-descent trajectories as the root of a nonlinear operator and uses parallel root-finding to update multiple optimizer steps simultaneously to cut wall-clock training time.

Problem

Wall-clock cost of training large neural networks is dominated by the long sequential chain of optimizer iterations.

Approach

Treats the GD (SGD/Adam) trajectory as the unique root of a nonlinear operator so that multiple GD steps can be solved for in parallel via root-finding, rather than computed one after another.

Experimental setup

Metrics: wall-clock time

Builds on: SGD Adam

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