← Authors

Itay Safran

9 papers · 2016–2025 · 3 conferences · across top CS/AI conferences

Achievements

🐝 Cross-Pollinator (8) 🌉 Interdisciplinary Bridge 🧭 Keyword Pioneer 🌍 Conference Polyglot (3) 🏃 Academic Marathon (9) 🐣 Hot Topic Early Bird ❓ The Questioner (2) 🔥 Unstoppable (7)

Conferences

COLT (4) ICML (3) NIPS (2)

Top co-authors

Ohad Shamir (6) Jason Lee (2) Ronen Eldan (1) Gal Vardi (1) Daniel Reichman (1) Paul Valiant (1)

Keywords

depth separation (2) relu network (2) stochastic gradient descent (2) finite-sum optimization (1) function approximation (1) convergence analysis (1) convex optimization (1) overspecified networks (1) initialization (1) loss landscape (1) gradient descent (1) strongly convex (1) optimization theory (1) condition number (1) convergence guarantee (1) generalization bound (1) convergence rate (1) gradient flow (1) lower bound (1) neural network optimization (1)

Papers

Depth Separations in Neural Networks: Separating the Dimension from the Accuracy COLT 2025
Optimization-Based Separations for Neural Networks COLT 2022
On the Effective Number of Linear Regions in Shallow Univariate ReLU Networks: Convergence Guarantees and Implicit Bias NIPS 2022
Random Shuffling Beats SGD Only After Many Epochs on Ill-Conditioned Problems NIPS 2021
How Good is SGD with Random Shuffling? COLT 2020
Depth Separations in Neural Networks: What is Actually Being Separated? COLT 2019
Spurious Local Minima are Common in Two-Layer ReLU Neural Networks ICML 2018
Depth-Width Tradeoffs in Approximating Natural Functions with Neural Networks ICML 2017
On the Quality of the Initial Basin in Overspecified Neural Networks ICML 2016