AI Paper Atlas

Research map / Optimization / Optimizers

ZO-AdaMM: Zeroth-Order Adaptive Momentum Method for Black-Box Optimization

NeurIPS 2019 · Optimization › Optimizers · approach: Zeroth-order training methods · 33 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use adaptive momentum (ZO-AdaMM) instead of Mahalanobis-free ZO-SMDZO-SMD · ImageNet universal black-box attack, 100 images, T=40000attack success rate: 79 -> 84Helped
use ZO-AdaMM instead of ZO-SMDZO-SMD · ImageNet universal black-box attack, 100 images, T=40000final squared l2 distortion: 47.36 -> 38.40Helped
use ZO-AdaMM instead of ZO-NES (Euclidean projected sign-based)ZO-NES · ImageNet universal black-box attack, 100 images, T=40000attack success rate: 74 -> 84Helped
use ZO-AdaMM instead of ZO-PSGDZO-PSGD · ImageNet universal black-box attack, 100 images, T=40000final squared l2 distortion: 49.92 -> 38.40Helped
use ZO-AdaMM instead of ZO-signSGD / ZO-SGD / ZO-SCDunconstrained per-image attack · ImageNet per-image attack, 100 imagesattack loss convergence and l2 distortionHelped

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