Simmering: Sufficient is better than optimal for training neural networks

Simmering implements a version of the sufficient training paradigm introduced in 10.1038/s41467-025-66983-3.

cs.LG cs.AI Neural Networks and Applications Computer Science Physics Medicine

View graph Source

Paper

Similar papers

Sufficient is better than optimal for training neural networksIrina Babayan, Hazhir Aliahmadi, Greg van Anders2025 · Nature Communications In Library

Langevin Dynamics with Continuous Tempering for Training Deep Neural NetworksNanyang Ye, Zhanxing Zhu, Rafał K. Mantiuk2017 · Neural Information Processing Systems · 21 citations In Library

Training Deep Neural Networks by optimizing over nonlocal paths in hyperparameter spaceVlad Pushkarov, Jonathan Efroni, M. Maksymenko et al.2019 · arXiv.org · 1 citation In Library

Toward Moderate Overparameterization: Global Convergence Guarantees for Training Shallow Neural NetworksSamet Oymak, Mahdi Soltanolkotabi2020 · IEEE Journal on Selected Areas in Information Theory In Library

Enhanced Self-Distillation Framework for Efficient Spiking Neural Network TrainingXiaochen Zhao, Chengting Yu, Kairong Yu et al.2025 · Neural Information Processing Systems · 0 citations In Library

Masked Training of Neural Networks with Partial GradientsAmirkeivan Mohtashami, Martin Jaggi, Sebastian U. Stich2021 · arXiv (Cornell University) · 30 citations In Library

Scheduled Runtime Refinement for Neural Networks: Periodic Pruning, Replay, and Distillation as an Operating RegimeRiaan De Beer2026 · Zenodo (CERN European Organization for Nuclear Research) In Library

Scheduled Runtime Refinement for Neural Networks: Periodic Pruning, Replay, and Distillation as an Operating RegimeRiaan De Beer2026 · Zenodo (CERN European Organization for Nuclear Research) In Library

Scheduled Runtime Refinement for Neural Networks: Periodic Pruning, Replay, and Distillation as an Operating RegimeRiaan De Beer2026 · Zenodo (CERN European Organization for Nuclear Research) In Library

Scheduled Runtime Refinement for Neural Networks: Periodic Pruning, Replay, and Distillation as an Operating RegimeRiaan De Beer2026 · Zenodo (CERN European Organization for Nuclear Research) In Library

Source of record

arXiv

44B

Welcome.

This panel sticks with you. Pick anything below and it opens right beside it, so you can dig through 60,000-plus records without ever losing your spot here.

Travel 44B

LibraryPapers, policy, standards, statute LabsEvery organization building AI ModelsIntelligence, price, and speed BenchmarksThe evaluation catalog IncidentsAI failures and harms SearchOne field across everything DashboardThe state of AI accountability in NY 44B RegistryThe Art. 44-B compliance portal