Research map / Optimization / Nonconvex and smooth optimization
NEON2: Finding Local Minima via First-Order Oracles
Zeyuan Allen-Zhu, Yuanzhi Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Replace Neon with Neon2 in the SGD local-minimum method | SGD · Online local minima | stochastic gradient complexity: eO(1/ε4 + 1/δ7) -> eO(1/ε4 + 1/(ε2δ3) + 1/δ5) | Improved |
| Replace Neon with Neon2 when converting Natasha2 | Natasha2 · Online local minima | stochastic gradient complexity: eO(1/ε3.25 + 1/(ε3δ) + 1/δ6) -> eO(1/ε3.25 + 1/(ε3δ) + 1/δ5) | Improved |
| Use Neon2det instead of Hessian-vector products for NC-search | CDHS · Offline local minima | full gradient complexity | No clear effect |
| Use Neon2 to convert stationary-point methods into local-minimum methods | SGD, GD, SCSG, and SVRG · Local minima | gradient complexity | No clear effect |
| Replace Hessian-vector products with gradient computations in NC-search | Neon2 · Online, deterministic, and finite-sum NC-search | gradient complexity | No clear effect |
Closest papers by meaning
- Natasha 2: Faster Non-Convex Optimization Than SGD (NeurIPS 2018 · similarity 0.88)
- Third-order Smoothness Helps: Faster Stochastic Optimization Algorithms for Finding Local Minima (NeurIPS 2018 · similarity 0.88)
- Simple and Optimal Stochastic Gradient Methods for Nonsmooth Nonconvex Optimization (JMLR 2022 · similarity 0.87)
- Balancing Gradient and Hessian Queries in Non-Convex Optimization (NeurIPS 2025 · similarity 0.86)
- A Study of First-Order Methods with a Deterministic Relative-Error Gradient Oracle (ICML 2024 · similarity 0.86)
- On The Complexity of First-Order Methods in Stochastic Bilevel Optimization (ICML 2024 · similarity 0.86)
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.
