JMLR 2025 papers on Optimization
51 of the 298 papers published in JMLR 2025 (17.1%) are on Optimization, compared with 14.6% at JMLR 2024. They fall into 4 topics.
See JMLR 2025 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Nonconvex and smooth optimization · 33 papers
- Near-Optimal Nonconvex-Strongly-Convex Bilevel Optimization with Fully First-Order Oracles (23 citations)
- Revisiting Gradient Normalization and Clipping for Nonconvex SGD under Heavy-Tailed Noise: Necessity, Sufficiency, and Acceleration (18 citations)
- Convergence Rates for Non-Log-Concave Sampling and Log-Partition Estimation (13 citations)
- Decentralized Bilevel Optimization: A Perspective from Transient Iteration Complexity (10 citations)
- Nonconvex Stochastic Bregman Proximal Gradient Method with Application to Deep Learning (10 citations)
- On the Convergence of Projected Policy Gradient for Any Constant Step Sizes (8 citations)
- Error bounds for particle gradient descent, and extensions of the log-Sobolev and Talagrand inequalities (7 citations)
- The ODE Method for Stochastic Approximation and Reinforcement Learning with Markovian Noise (6 citations)
- A Decentralized Proximal Gradient Tracking Algorithm for Composite Optimization on Riemannian Manifolds (4 citations)
- Accelerating optimization over the space of probability measures (4 citations)
- Distributed Stochastic Bilevel Optimization: Improved Complexity and Heterogeneity Analysis (4 citations)
- Riemannian Bilevel Optimization (4 citations)
and 21 more in the interactive map.
Implicit bias and generalization · 12 papers
- Early Alignment in Two-Layer Networks Training is a Two-Edged Sword (14 citations)
- Good regularity creates large learning rate implicit biases: edge of stability, balancing, and catapult (14 citations)
- Characterizing Dynamical Stability of Stochastic Gradient Descent in Overparameterized Learning (4 citations)
- DRM Revisited: A Complete Error Analysis (3 citations)
- General Loss Functions Lead to (Approximate) Interpolation in High Dimensions (3 citations)
- On the Utility of Equal Batch Sizes for Inference in Stochastic Gradient Descent (3 citations)
- Stabilizing Sharpness-Aware Minimization Through A Simple Renormalization Strategy (3 citations)
- Towards Understanding Gradient Flow Dynamics of Homogeneous Neural Networks Beyond the Origin (2 citations)
- The Effect of SGD Batch Size on Autoencoder Learning: Sparsity, Sharpness, and Feature Learning (1 citation)
- Almost Sure Convergence of Dropout Algorithms for Neural Networks
- Enhanced Feature Learning via Regularisation: Integrating Neural Networks and Kernel Methods
- Random Pruning Over-parameterized Neural Networks Can Improve Generalization: A Training Dynamics Analysis
