Double Descent

Ask about this topic

Overview

Double descent studies why test error can rise near the interpolation threshold, then fall as models become more overparameterized. Researchers use random-matrix and asymptotic analyses of ridgeless regression and kernel methods, alongside experiments that vary model size in neural networks, boosting, and parameter-efficient fine-tuning. Work also examines how implicit or explicit regularization, data noise, and signal alignment shape these risk curves. The central open tension is that overparameterization can improve generalization in some settings but not others, so the phenomenon is not a guarantee of better performance.

Latest papers

CardsList

20 papers1 paper in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.

  1. LoRA's Second Descent Extends Beyond Parameter Parity

    Oct 3, 2026Yueran MaDouble DescentLow-Rank Adaptation

    The University of Queensland Brisbane, Australia

  2. Double Descent and Malign Overfitting in Diffusion Models

    Sep 22, 2026Raphaël Urfin, Tony Bonnaire, Giulio Biroli +1Double DescentBenign Overfitting

    Laboratoire de Physique de l’École normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris Cité, F-75005 Paris, France · Université Paris-Saclay, CNRS, Institut d’Astrophysique Spatiale, 91405 Orsay, France · Department of Computing Sciences, Bocconi University, Milano, Italy

  3. Double descent is the principle of least action

    Sep 16, 2026Congzhou M ShaDouble DescentStatistical Physics of Learning

    Penn Medicine Doylestown Hospital, 595 W State St, Doylestown, PA 18901, United States

  4. Landau theory of quenched criticality in linear in-context learning

    Aug 28, 2026Daesik Kim, Sumin Choi, Hyojae Jeon +1In-Context LearningDouble Descent

    Department of Physics, Sungkyunkwan University, Suwon 16419, South Korea · Center for Quantum Dynamics of Angular Momentum, Pohang University of Science and Technology, Pohang 37673, Korea

  5. (Mis)Understanding Benign Overfitting in Equity Return Prediction

    Aug 24, 2026Hui Guo, Jiawei Huang, Runze Li +1Benign OverfittingQuantitative Finance

    Carl H. Lindner College of Business, University of Cincinnati · Católica Lisbon School of Business and Economics, Universidade Católica Portuguesa · Department of Statistics, Pennsylvania State University

  6. High-dimensional ridgeless least squares interpolation under spiked covariance structures

    Aug 7, 2026Zhijun Liu, Dandan JiangDouble DescentBenign Overfitting

    College of Sciences, Northeastern University, China. · School of Mathematics and Statistics, Xi’an Jiao Tong University, China.

  7. Double Descent in Gradient Boosting Decision Trees via Split-Candidate Scaling

    Aug 4, 2026Ryuichi KanohDouble DescentGradient-Boosted Decision Trees

    1The University of Electro-Communications · 2National Institute of Informatics

  8. Origins and mitigation of double descent in reduced order modeling

    Jul 29, 2026Andrei A. Klishin, J. Nathan Kutz, Krithika ManoharDouble DescentSparse Recovery

    Department of Mechanical Engineering, University of Hawai‘ i at M¯anoa, Honolulu, HI · AI Institute in Dynamic Systems, University of Washington, Seattle, WA 98195, USA · Departments of Applied Mathematics and Electrical and Computer Engineering, University of Washington, Seattle, WA 98195, USA +1

  9. The Zero Pattern of a Design Matrix Drives Multiple Descent in Over-parameterized Regression

    Jul 27, 2026Kevin Han Huang, Haoyu Ye, Somak Laha +1Double DescentLinear Regression

    University of Warwick · Harvard University

  10. Cautious optimism for deep parameterized quantum circuits

    Jul 23, 2026Marie Kempkes, Elies Gil-Fuster, Carlos Bravo-Prieto +5Double DescentNeural Network Generalization

    Leiden University, Niels Bohrweg 1, 2333 CA Leiden, Netherlands · Volkswagen Group Innovation, Berliner Ring 2, 38440 Wolfsburg, Germany · Dahlem Center for Complex Quantum Systems, Freie Universität Berlin, 14195 Berlin, Germany +4

  11. Broken Ergodicity and the Violation of the Fluctuation-Dissipation Theorem Lead to Generalization Beyond Overfitting in Machine Learning

    Jul 5, 2026Chan Li, Nigel GoldenfeldDynamical Mean-Field TheoryStatistical Physics of Learning

    Department of Physics, University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093, USA. · Halıcıoğlu Data Science Institute, University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093, USA

  12. Benign Overfitting Does Not Occur in Diffusion Models

    Jul 2, 2026Tyler Farghly, Benjamin Dupuis, Alain Durmus +1Double DescentNeural Network Generalization

    1INRIA - École Normale Supérieure - PSL Research University - CNRS, France · 2 École Polytechnique - CMAP - IP Paris, France

  13. Two Speeds of Learning: A Representation-Readout Decomposition of Grokking and Double Descent

    May 26, 2026Chi-Ning Chou, Oscar Uzdelewicz, Neng-Chun Chiu +2Double DescentNeural Network Generalization

    Center for Computational Neuroscience, Flatiron Institute, New York, NY, USA. · Department of Physics, Harvard University, Cambridge, MA, USA. · Kempner Institute, Harvard University, Allston, MA, USA. +1

  14. Asymmetric Scaling Laws from Sparse Features

    May 22, 2026John Sous, Michael WinerDouble DescentActivation Sparsity

    Department of Applied Physics, Yale University, New Haven, Connecticut 06511, USA · Energy Sciences Institute, Yale University, West Haven, Connecticut 06516, USA · Institute for Advanced Study, Princeton, NJ 08540, USA +1

  15. Large Dimensional Kernel Ridge Regression: Extending to Product Kernels

    May 14, 2026Yang Zhou, Yicheng Li, Yuqian Cheng +1Generalization BoundsKernel Ridge Regression

    Department of Statistics and Data Science Tsinghua University Beijing, 100084, China · Department of Mathematical Science Tsinghua University Beijing, 100084, China

  16. Double descent for least-squares interpolation on contaminated data: A simulation study

    Apr 15, 2026Tino WernerDouble DescentLinear Regression

    Institute for Mathematics, Carl von Ossietzky University Oldenburg, Carl-von-Ossietzky-Straße 9-11, 26129 Oldenburg, Lower Saxony, Germany

  17. Asymptotic Behavior of Multi--Task Learning: Implicit Regularization and Double Descent Effects

    Mar 5, 2026Ayed M. Alrashdi, Oussama Dhifallah, Houssem SifaouDouble DescentNeural Network Generalization

    Department of Electrical Engineering, College of Engineering, University of Ha’il, P.O. Box 2440, Ha’il, 81441, Saudi Arabia · John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138, USA · Department of Electrical and Electronic Engineering, King’s College London – Strand, London WC2R 2LS – London, UK

  18. Robust Predictive Uncertainty and Double Descent in Contaminated Bayesian Random Features

    Feb 22, 2026Michele Caprio, Katerina Papagiannouli, Siu Lun Chau +1Random Feature MethodsBayesian Inference

    The University of Manchester, UK · University of Pisa, Italy · Nanyang Technological University, Singapore +1

  19. The Mathematical Theory of Double Descent as Shrinkage, not Overfitting: Matching Bounds for Path-Regularized Multilayer Neural Networks

    Mar 3, 2025Hao YuDouble DescentNeural Network Generalization

    Department of Automation, Tsinghua University, Beijing, 100086, China

  20. A Farewell to the Bias-Variance Tradeoff? An Overview of the Theory of Overparameterized Machine Learning

    Sep 6, 2021Yehuda Dar, Vidya Muthukumar, Richard G. BaraniukDouble DescentNeural Network Generalization

    Faculty of Computer and Information Science, Ben-Gurion University · Schools of Electrical and Computer Engineering, and Industrial and Systems Engineering, Georgia Institute of Technology · Electrical and Computer Engineering Department, Rice University