{"total":162,"rows":[{"id":"2101.12176","title":"On the Origin of Implicit Regularization in Stochastic Gradient Descent","created":"2021-01-28","cited_by_count":184,"area":null},{"id":"2203.08065","title":"Surrogate Gap Minimization Improves Sharpness-Aware Training","created":"2022-03-15","cited_by_count":172,"area":null},{"id":"2102.03526","title":"Open-World Semi-Supervised Learning","created":"2021-02-06","cited_by_count":165,"area":null},{"id":"2209.15594","title":"Self-Stabilization: The Implicit Bias of Gradient Descent at the Edge of Stability","created":"2022-09-30","cited_by_count":155,"area":null},{"id":"2205.01445","title":"High-dimensional Asymptotics of Feature Learning: How One Gradient Step Improves the Representation","created":"2022-05-03","cited_by_count":150,"area":null},{"id":"2106.06530","title":"Label Noise SGD Provably Prefers Flat Global Minimizers","created":"2021-06-11","cited_by_count":121,"area":null},{"id":"2110.06914","title":"What Happens after SGD Reaches Zero Loss? --A Mathematical Framework","created":"2021-10-13","cited_by_count":107,"area":null},{"id":"2210.09221","title":"Vision Transformers provably learn spatial structure","created":"2022-10-13","cited_by_count":93,"area":null},{"id":"2302.07011","title":"A Modern Look at the Relationship between Sharpness and Generalization","created":"2023-02-14","cited_by_count":92,"area":null},{"id":"2202.06526","title":"Benign Overfitting in Two-layer Convolutional Neural Networks","created":"2022-02-14","cited_by_count":91,"area":null},{"id":"2202.08345","title":"Learning Smooth Neural Functions via Lipschitz Regularization","created":"2022-02-16","cited_by_count":86,"area":null},{"id":"2308.16898","title":"Transformers as Support Vector Machines","created":"2023-08-31","cited_by_count":74,"area":null},{"id":"2102.12470","title":"On the Validity of Modeling SGD with Stochastic Differential Equations (SDEs)","created":"2021-02-24","cited_by_count":73,"area":null},{"id":"2310.04415","title":"Why Do We Need Weight Decay in Modern Deep Learning?","created":"2023-10-06","cited_by_count":72,"area":null},{"id":"2210.05337","title":"SGD with Large Step Sizes Learns Sparse Features","created":"2022-10-11","cited_by_count":68,"area":null},{"id":"2112.04716","title":"DR3: Value-Based Deep Reinforcement Learning Requires Explicit Regularization","created":"2021-12-09","cited_by_count":66,"area":"offlinerl"},{"id":"2109.14119","title":"Stochastic Training is Not Necessary for Generalization","created":"2021-09-29","cited_by_count":63,"area":null},{"id":"2306.13596","title":"Max-Margin Token Selection in Attention Mechanism","created":"2023-06-23","cited_by_count":58,"area":null},{"id":"2404.04454","title":"Implicit Bias of AdamW: $\\ell_\\infty$ Norm Constrained Optimization","created":"2024-04-05","cited_by_count":58,"area":null},{"id":"2212.07469","title":"Learning threshold neurons via the \"edge of stability\"","created":"2022-12-14","cited_by_count":51,"area":null},{"id":"2105.02716","title":"Noether's Learning Dynamics: Role of Symmetry Breaking in Neural Networks","created":"2021-05-06","cited_by_count":50,"area":null},{"id":"2110.03677","title":"Large Learning Rate Tames Homogeneity: Convergence and Balancing Effect","created":"2021-10-07","cited_by_count":50,"area":null},{"id":"2206.01717","title":"A Theoretical Analysis on Feature Learning in Neural Networks: Emergence from Inputs and Advantage over Fixed Features","created":"2022-06-03","cited_by_count":49,"area":null},{"id":"2106.00042","title":"A study on the plasticity of neural networks","created":"2021-05-31","cited_by_count":44,"area":null},{"id":"2303.15725","title":"Solving Regularized Exp, Cosh and Sinh Regression Problems","created":"2023-03-28","cited_by_count":44,"area":null},{"id":"2307.06440","title":"No Train No Gain: Revisiting Efficient Training Algorithms For Transformer-based Language Models","created":"2023-07-12","cited_by_count":41,"area":null},{"id":"2207.05931","title":"Towards understanding how momentum improves generalization in deep learning","created":"2022-07-13","cited_by_count":40,"area":null},{"id":"2207.04036","title":"Implicit Bias of Gradient Descent on Reparametrized Models: On Equivalence to Mirror Descent","created":"2022-07-08","cited_by_count":39,"area":null},{"id":"2205.06226","title":"The Mechanism of Prediction Head in Non-contrastive Self-supervised Learning","created":"2022-05-12","cited_by_count":38,"area":null},{"id":"2210.15031","title":"Characterizing Datapoints via Second-Split Forgetting","created":"2022-10-26","cited_by_count":37,"area":null},{"id":"2302.08982","title":"(S)GD over Diagonal Linear Networks: Implicit Regularisation, Large Stepsizes and Edge of Stability","created":"2023-02-17","cited_by_count":36,"area":null},{"id":"2107.06608","title":"Continuous vs. Discrete Optimization of Deep Neural Networks","created":"2021-07-14","cited_by_count":34,"area":null},{"id":"2303.04145","title":"Benign Overfitting for Two-layer ReLU Convolutional Neural Networks","created":"2023-03-07","cited_by_count":34,"area":null},{"id":"2403.08081","title":"Mechanics of Next Token Prediction with Self-Attention","created":"2024-03-12","cited_by_count":34,"area":null},{"id":"2105.13343","title":"Drawing Multiple Augmentation Samples Per Image During Training Efficiently Decreases Test Error","created":"2021-05-27","cited_by_count":33,"area":null},{"id":"2211.06506","title":"Spectral Evolution and Invariance in Linear-width Neural Networks","created":"2022-11-11","cited_by_count":30,"area":null},{"id":"2211.12491","title":"ModelDiff: A Framework for Comparing Learning Algorithms","created":"2022-11-22","cited_by_count":30,"area":null},{"id":"2308.12553","title":"Don't blame Dataset Shift! Shortcut Learning due to Gradients and Cross Entropy","created":"2023-08-24","cited_by_count":30,"area":null},{"id":"2309.07418","title":"A Fast Optimization View: Reformulating Single Layer Attention in LLM Based on Tensor and SVM Trick, and Solving It in Matrix Multiplication Time","created":"2023-09-14","cited_by_count":30,"area":null},{"id":"2311.14495","title":"StableSSM: Alleviating the Curse of Memory in State-space Models through Stable Reparameterization","created":"2023-11-24","cited_by_count":30,"area":null},{"id":"2311.17971","title":"GeoDream: Disentangling 2D and Geometric Priors for High-Fidelity and Consistent 3D Generation","created":"2023-11-29","cited_by_count":30,"area":"diffusion"},{"id":"2103.17182","title":"Positive-Negative Momentum: Manipulating Stochastic Gradient Noise to Improve Generalization","created":"2021-03-31","cited_by_count":28,"area":null},{"id":"2110.11222","title":"Is High Variance Unavoidable in RL? A Case Study in Continuous Control","created":"2021-10-21","cited_by_count":28,"area":null},{"id":"2310.07269","title":"Why Does Sharpness-Aware Minimization Generalize Better Than SGD?","created":"2023-10-11","cited_by_count":28,"area":null},{"id":"2206.03093","title":"Beyond spectral gap: The role of the topology in decentralized learning","created":"2022-06-07","cited_by_count":27,"area":null},{"id":"2212.13881","title":"Mechanism of feature learning in deep fully connected networks and kernel machines that recursively learn features","created":"2022-12-28","cited_by_count":25,"area":null},{"id":"2509.20758","title":"SFT Doesn't Always Hurt General Capabilities: Revisiting Domain-Specific Fine-Tuning in LLMs","created":"2025-09-25","cited_by_count":24,"area":null},{"id":"2310.07999","title":"LEMON: Lossless model expansion","created":"2023-10-12","cited_by_count":23,"area":"lr"},{"id":"2402.17457","title":"Super Consistency of Neural Network Landscapes and Learning Rate Transfer","created":"2024-02-27","cited_by_count":23,"area":null},{"id":"2307.06966","title":"Layer-wise Linear Mode Connectivity","created":"2023-07-13","cited_by_count":22,"area":null},{"id":"2503.12811","title":"A Multi-Power Law for Loss Curve Prediction Across Learning Rate Schedules","created":"2025-03-17","cited_by_count":22,"area":null},{"id":"2402.18137","title":"DecisionNCE: Embodied Multimodal Representations via Implicit Preference Learning","created":"2024-02-28","cited_by_count":20,"area":null},{"id":"2101.09868","title":"CPT: Efficient Deep Neural Network Training via Cyclic Precision","created":"2021-01-25","cited_by_count":19,"area":null},{"id":"2209.03695","title":"Training Scale-Invariant Neural Networks on the Sphere Can Happen in Three Regimes","created":"2022-09-08","cited_by_count":19,"area":null},{"id":"2310.19360","title":"Balance, Imbalance, and Rebalance: Understanding Robust Overfitting from a Minimax Game Perspective","created":"2023-10-30","cited_by_count":18,"area":null},{"id":"2101.01152","title":"Provable Generalization of SGD-trained Neural Networks of Any Width in the Presence of Adversarial Label Noise","created":"2021-01-04","cited_by_count":17,"area":null},{"id":"2502.03435","title":"Taking a Big Step: Large Learning Rates in Denoising Score Matching Prevent Memorization","created":"2025-02-05","cited_by_count":17,"area":"diffusion"},{"id":"2305.12133","title":"Loss Spike in Training Neural Networks","created":"2023-05-20","cited_by_count":16,"area":null},{"id":"2307.06887","title":"Provable Multi-Task Representation Learning by Two-Layer ReLU Neural Networks","created":"2023-07-13","cited_by_count":16,"area":null},{"id":"2102.12430","title":"Noisy Gradient Descent Converges to Flat Minima for Nonconvex Matrix Factorization","created":"2021-02-24","cited_by_count":15,"area":null},{"id":"2306.17499","title":"The Implicit Bias of Minima Stability in Multivariate Shallow ReLU Networks","created":"2023-06-30","cited_by_count":15,"area":null},{"id":"2406.03880","title":"Memorization in deep learning: A survey","created":"2024-06-06","cited_by_count":15,"area":null},{"id":"2310.11428","title":"Butterfly Effects of SGD Noise: Error Amplification in Behavior Cloning and Autoregression","created":"2023-10-17","cited_by_count":14,"area":null},{"id":"2103.01338","title":"Acceleration via Fractal Learning Rate Schedules","created":"2021-03-01","cited_by_count":13,"area":null},{"id":"2311.04163","title":"Outliers with Opposing Signals Have an Outsized Effect on Neural Network Optimization","created":"2023-11-07","cited_by_count":13,"area":null},{"id":"2111.00843","title":"How I Learned to Stop Worrying and Love Retraining","created":"2021-11-01","cited_by_count":12,"area":null},{"id":"2206.10011","title":"When Does Re-initialization Work?","created":"2022-06-20","cited_by_count":12,"area":null},{"id":"2210.10737","title":"RSC: Accelerating Graph Neural Networks Training via Randomized Sparse Computations","created":"2022-10-19","cited_by_count":12,"area":null},{"id":"2103.12682","title":"How to decay your learning rate","created":"2021-03-23","cited_by_count":11,"area":null},{"id":"2212.07295","title":"Maximal Initial Learning Rates in Deep ReLU Networks","created":"2022-12-14","cited_by_count":11,"area":null},{"id":"2302.01952","title":"On a continuous time model of gradient descent dynamics and instability in deep learning","created":"2023-02-03","cited_by_count":11,"area":null},{"id":"2205.15142","title":"Special Properties of Gradient Descent with Large Learning Rates","created":"2022-05-30","cited_by_count":10,"area":null},{"id":"2310.17588","title":"PAC-tuning:Fine-tuning Pretrained Language Models with PAC-driven Perturbed Gradient Descent","created":"2023-10-26","cited_by_count":10,"area":null},{"id":"2605.02105","title":"Sharpness-Aware Pretraining Mitigates Catastrophic Forgetting","created":"2026-05-04","cited_by_count":10,"area":null},{"id":"2205.07320","title":"Analyzing Lottery Ticket Hypothesis from PAC-Bayesian Theory Perspective","created":"2022-05-15","cited_by_count":9,"area":null},{"id":"2212.08378","title":"Feature Dropout: Revisiting the Role of Augmentations in Contrastive Learning","created":"2022-12-16","cited_by_count":9,"area":null},{"id":"2310.11291","title":"An Automatic Learning Rate Schedule Algorithm for Achieving Faster Convergence and Steeper Descent","created":"2023-10-17","cited_by_count":9,"area":null},{"id":"2103.13462","title":"Why Do Local Methods Solve Nonconvex Problems?","created":"2021-03-24","cited_by_count":8,"area":null},{"id":"2201.03789","title":"Partial Model Averaging in Federated Learning: Performance Guarantees and Benefits","created":"2022-01-11","cited_by_count":8,"area":null},{"id":"2202.13733","title":"On the Benefits of Large Learning Rates for Kernel Methods","created":"2022-02-28","cited_by_count":8,"area":null},{"id":"2301.07737","title":"Catapult Dynamics and Phase Transitions in Quadratic Nets","created":"2023-01-18","cited_by_count":8,"area":null},{"id":"2512.25034","title":"Generative Classifiers Avoid Shortcut Solutions","created":"2025-12-31","cited_by_count":8,"area":null},{"id":"2104.00008","title":"Why is AI hard and Physics simple?","created":"2021-03-31","cited_by_count":7,"area":null},{"id":"2110.01543","title":"Stochastic Anderson Mixing for Nonconvex Stochastic Optimization","created":"2021-10-04","cited_by_count":7,"area":null},{"id":"2210.05371","title":"On skip connections and normalisation layers in deep optimisation","created":"2022-10-10","cited_by_count":7,"area":null},{"id":"2108.06808","title":"Implicit Regularization of Bregman Proximal Point Algorithm and Mirror Descent on Separable Data","created":"2021-08-15","cited_by_count":6,"area":null},{"id":"2201.09209","title":"Weight Expansion: A New Perspective on Dropout and Generalization","created":"2022-01-23","cited_by_count":6,"area":null},{"id":"2206.07892","title":"Max-Margin Works while Large Margin Fails: Generalization without Uniform Convergence","created":"2022-06-16","cited_by_count":6,"area":null},{"id":"2401.11600","title":"Understanding the Generalization Benefits of Late Learning Rate Decay","created":"2024-01-21","cited_by_count":6,"area":null},{"id":"2403.07688","title":"Maxwell's Demon at Work: Efficient Pruning by Leveraging Saturation of Neurons","created":"2024-03-12","cited_by_count":6,"area":null},{"id":"2410.22113","title":"Where Do Large Learning Rates Lead Us?","created":"2024-10-29","cited_by_count":6,"area":null},{"id":"2502.02132","title":"How Memory in Optimization Algorithms Implicitly Modifies the Loss","created":"2025-02-04","cited_by_count":6,"area":null},{"id":"2505.15270","title":"Scaling Diffusion Transformers Efficiently via $\\mu$P","created":"2025-05-21","cited_by_count":6,"area":"diffusion"},{"id":"2604.13627","title":"(How) Learning Rates Regulate Catastrophic Overtraining","created":"2026-04-15","cited_by_count":6,"area":null},{"id":"2604.20446","title":"The Origin of Edge of Stability","created":"2026-04-22","cited_by_count":6,"area":null},{"id":"2106.10771","title":"Multirate Training of Neural Networks","created":"2021-06-20","cited_by_count":5,"area":null},{"id":"2203.04668","title":"Towards Inadequately Pre-trained Models in Transfer Learning","created":"2022-03-09","cited_by_count":5,"area":null},{"id":"2209.09658","title":"Lazy vs hasty: linearization in deep networks impacts learning schedule based on example difficulty","created":"2022-09-19","cited_by_count":5,"area":null},{"id":"2304.12579","title":"Learning Trajectories are Generalization Indicators","created":"2023-04-25","cited_by_count":5,"area":null},{"id":"2306.04793","title":"On the Joint Interaction of Models, Data, and Features","created":"2023-06-07","cited_by_count":5,"area":null}],"next_cursor":"100"}