{"total":160,"rows":[{"id":"2102.11535","title":"Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective","created":"2021-02-23","cited_by_count":169,"area":null},{"id":"2102.02611","title":"CKConv: Continuous Kernel Convolution For Sequential Data","created":"2021-02-04","cited_by_count":120,"area":null},{"id":"2210.17409","title":"Deep Model Reassembly","created":"2022-10-24","cited_by_count":78,"area":null},{"id":"2102.01063","title":"Zen-NAS: A Zero-Shot NAS for High-Performance Deep Image Recognition","created":"2021-02-01","cited_by_count":73,"area":null},{"id":"2103.05127","title":"Model Complexity of Deep Learning: A Survey","created":"2021-03-08","cited_by_count":65,"area":null},{"id":"2201.10355","title":"Neural Architecture Search for Spiking Neural Networks","created":"2022-01-23","cited_by_count":53,"area":null},{"id":"2211.14400","title":"Optimal Approximation Rates for Deep ReLU Neural Networks on Sobolev and Besov Spaces","created":"2022-11-25","cited_by_count":51,"area":null},{"id":"2402.15555","title":"Deep Networks Always Grok and Here is Why","created":"2024-02-23","cited_by_count":51,"area":null},{"id":"2107.01562","title":"Random Neural Networks in the Infinite Width Limit as Gaussian Processes","created":"2021-07-04","cited_by_count":44,"area":null},{"id":"2104.04413","title":"Provable Repair of Deep Neural Networks","created":"2021-04-09","cited_by_count":42,"area":null},{"id":"2406.06158","title":"Get rich quick: exact solutions reveal how unbalanced initializations promote rapid feature learning","created":"2024-06-10","cited_by_count":42,"area":null},{"id":"2203.04466","title":"The Combinatorial Brain Surgeon: Pruning Weights That Cancel One Another in Neural Networks","created":"2022-03-09","cited_by_count":41,"area":null},{"id":"2305.00241","title":"When Deep Learning Meets Polyhedral Theory: A Survey","created":"2023-04-29","cited_by_count":39,"area":null},{"id":"2104.08135","title":"Sharp bounds for the number of regions of maxout networks and vertices of Minkowski sums","created":"2021-04-16","cited_by_count":37,"area":null},{"id":"2305.19753","title":"The Tunnel Effect: Building Data Representations in Deep Neural Networks","created":"2023-05-31","cited_by_count":34,"area":null},{"id":"2302.12828","title":"SplineCam: Exact Visualization and Characterization of Deep Network Geometry and Decision Boundaries","created":"2023-02-24","cited_by_count":31,"area":null},{"id":"2403.19232","title":"AZ-NAS: Assembling Zero-Cost Proxies for Network Architecture Search","created":"2024-03-28","cited_by_count":28,"area":null},{"id":"2210.07236","title":"Improved Bounds on Neural Complexity for Representing Piecewise Linear Functions","created":"2022-10-13","cited_by_count":26,"area":null},{"id":"2106.13361","title":"Multifidelity Modeling for Physics-Informed Neural Networks (PINNs)","created":"2021-06-25","cited_by_count":25,"area":null},{"id":"2207.07696","title":"Algorithmic Determination of the Combinatorial Structure of the Linear Regions of ReLU Neural Networks","created":"2022-07-15","cited_by_count":22,"area":null},{"id":"2206.08615","title":"On the Number of Regions of Piecewise Linear Neural Networks","created":"2022-06-17","cited_by_count":21,"area":null},{"id":"2307.01998","title":"Zero-Shot Neural Architecture Search: Challenges, Solutions, and Opportunities","created":"2023-07-05","cited_by_count":21,"area":null},{"id":"2307.06092","title":"Quantitative CLTs in Deep Neural Networks","created":"2023-07-12","cited_by_count":20,"area":null},{"id":"2102.07804","title":"Scaling Up Exact Neural Network Compression by ReLU Stability","created":"2021-02-15","cited_by_count":19,"area":null},{"id":"2102.01621","title":"Depth separation beyond radial functions","created":"2021-02-02","cited_by_count":18,"area":null},{"id":"2102.10492","title":"Deep ReLU Networks Preserve Expected Length","created":"2021-02-21","cited_by_count":17,"area":null},{"id":"2111.13336","title":"MAE-DET: Revisiting Maximum Entropy Principle in Zero-Shot NAS for Efficient Object Detection","created":"2021-11-26","cited_by_count":16,"area":null},{"id":"2107.00379","title":"On the Expected Complexity of Maxout Networks","created":"2021-07-01","cited_by_count":15,"area":null},{"id":"2204.01368","title":"Training Fully Connected Neural Networks is $\\exists\\mathbb{R}$-Complete","created":"2022-04-04","cited_by_count":15,"area":null},{"id":"2306.07212","title":"Polyhedral Complex Extraction from ReLU Networks using Edge Subdivision","created":"2023-06-12","cited_by_count":14,"area":null},{"id":"2310.05918","title":"Grokking as Compression: A Nonlinear Complexity Perspective","created":"2023-10-09","cited_by_count":14,"area":null},{"id":"2405.05369","title":"Model Reconstruction Using Counterfactual Explanations: A Perspective From Polytope Theory","created":"2024-05-08","cited_by_count":14,"area":null},{"id":"2202.11749","title":"Are All Linear Regions Created Equal?","created":"2022-02-23","cited_by_count":13,"area":null},{"id":"2301.11701","title":"TransNet: Transferable Neural Networks for Partial Differential Equations","created":"2023-01-27","cited_by_count":13,"area":null},{"id":"2102.06635","title":"ReLU Neural Networks of Polynomial Size for Exact Maximum Flow Computation","created":"2021-02-12","cited_by_count":12,"area":null},{"id":"2106.03743","title":"Proxy-Normalizing Activations to Match Batch Normalization while Removing Batch Dependence","created":"2021-06-07","cited_by_count":12,"area":null},{"id":"2210.16114","title":"Towards Reliable Neural Specifications","created":"2022-10-28","cited_by_count":12,"area":null},{"id":"2301.00008","title":"Effects of Data Geometry in Early Deep Learning","created":"2022-12-29","cited_by_count":12,"area":null},{"id":"2307.09212","title":"How Many Neurons Does it Take to Approximate the Maximum?","created":"2023-07-18","cited_by_count":12,"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":"2301.12309","title":"On the Lipschitz Constant of Deep Networks and Double Descent","created":"2023-01-28","cited_by_count":11,"area":null},{"id":"2306.08805","title":"Exact Count of Boundary Pieces of ReLU Classifiers: Towards the Proper Complexity Measure for Classification","created":"2023-06-15","cited_by_count":11,"area":null},{"id":"2101.03263","title":"SyReNN: A Tool for Analyzing Deep Neural Networks","created":"2021-01-09","cited_by_count":10,"area":null},{"id":"2211.08404","title":"Non-Linear Coordination Graphs","created":"2022-10-26","cited_by_count":10,"area":null},{"id":"2401.03451","title":"Optimization Over Trained Neural Networks: Taking a Relaxing Walk","created":"2024-01-07","cited_by_count":10,"area":null},{"id":"2206.00228","title":"Lower and Upper Bounds for Numbers of Linear Regions of Graph Convolutional Networks","created":"2022-06-01","cited_by_count":9,"area":null},{"id":"2405.19816","title":"Growing Tiny Networks: Spotting Expressivity Bottlenecks and Fixing Them Optimally","created":"2024-05-30","cited_by_count":9,"area":null},{"id":"2412.18283","title":"On the Local Complexity of Linear Regions in Deep ReLU Networks","created":"2024-12-24","cited_by_count":9,"area":null},{"id":"2506.08001","title":"Reparameterized LLM Training via Orthogonal Equivalence Transformation","created":"2025-06-09","cited_by_count":9,"area":null},{"id":"2210.05742","title":"Curved Representation Space of Vision Transformers","created":"2022-10-11","cited_by_count":8,"area":null},{"id":"2210.12001","title":"When Expressivity Meets Trainability: Fewer than $n$ Neurons Can Work","created":"2022-10-21","cited_by_count":8,"area":null},{"id":"2301.07966","title":"Getting Away with More Network Pruning: From Sparsity to Geometry and Linear Regions","created":"2023-01-19","cited_by_count":8,"area":null},{"id":"2102.13233","title":"Spurious Local Minima Are Common for Deep Neural Networks with Piecewise Linear Activations","created":"2021-02-25","cited_by_count":7,"area":null},{"id":"2103.17174","title":"Using activation histograms to bound the number of affine regions in ReLU feed-forward neural networks","created":"2021-03-31","cited_by_count":7,"area":null},{"id":"2108.11939","title":"Understanding and Accelerating Neural Architecture Search with Training-Free and Theory-Grounded Metrics","created":"2021-08-26","cited_by_count":7,"area":null},{"id":"2110.15168","title":"Labeled sample compression schemes for complexes of oriented matroids","created":"2021-10-28","cited_by_count":7,"area":null},{"id":"2206.09149","title":"Piecewise Linear Neural Networks and Deep Learning","created":"2022-06-18","cited_by_count":7,"area":null},{"id":"2210.13611","title":"Understanding the Evolution of Linear Regions in Deep Reinforcement Learning","created":"2022-10-24","cited_by_count":7,"area":null},{"id":"2305.17154","title":"On convex decision regions in deep network representations","created":"2023-05-26","cited_by_count":7,"area":null},{"id":"2108.13637","title":"When are Deep Networks really better than Decision Forests at small sample sizes, and how?","created":"2021-08-31","cited_by_count":6,"area":null},{"id":"2210.08339","title":"Reachable Polyhedral Marching (RPM): An Exact Analysis Tool for Deep-Learned Control Systems","created":"2022-10-15","cited_by_count":6,"area":null},{"id":"2101.02338","title":"Max-Affine Spline Insights Into Deep Network Pruning","created":"2021-01-07","cited_by_count":5,"area":null},{"id":"2106.04186","title":"What training reveals about neural network complexity","created":"2021-06-08","cited_by_count":5,"area":null},{"id":"2209.13394","title":"Magnitude and Angle Dynamics in Training Single ReLU Neurons","created":"2022-09-27","cited_by_count":5,"area":null},{"id":"2210.05546","title":"What does a deep neural network confidently perceive? The effective dimension of high certainty class manifolds and their low confidence boundaries","created":"2022-10-11","cited_by_count":5,"area":null},{"id":"2305.09145","title":"Deep ReLU Networks Have Surprisingly Simple Polytopes","created":"2023-05-16","cited_by_count":5,"area":null},{"id":"2305.15912","title":"Neural Characteristic Activation Analysis and Geometric Parameterization for ReLU Networks","created":"2023-05-25","cited_by_count":5,"area":null},{"id":"2409.17113","title":"Characterizing stable regions in the residual stream of LLMs","created":"2024-09-25","cited_by_count":5,"area":null},{"id":"2505.07300","title":"L-SWAG: Layer-Sample Wise Activation with Gradients information for Zero-Shot NAS on Vision Transformers","created":"2025-05-12","cited_by_count":5,"area":null},{"id":"2101.04243","title":"A Convergence Theory Towards Practical Over-parameterized Deep Neural Networks","created":"2021-01-12","cited_by_count":4,"area":null},{"id":"2104.00219","title":"Fast Jacobian-Vector Product for Deep Networks","created":"2021-04-01","cited_by_count":4,"area":null},{"id":"2106.12074","title":"Reachability Analysis of Convolutional Neural Networks","created":"2021-06-22","cited_by_count":4,"area":null},{"id":"2112.04014","title":"Unsupervised Representation Learning via Neural Activation Coding","created":"2021-12-07","cited_by_count":4,"area":null},{"id":"2112.06209","title":"Measuring Complexity of Learning Schemes Using Hessian-Schatten Total Variation","created":"2021-12-12","cited_by_count":4,"area":null},{"id":"2203.13779","title":"Origins of Low-dimensional Adversarial Perturbations","created":"2022-03-25","cited_by_count":4,"area":null},{"id":"2402.11168","title":"Trust Regions for Explanations via Black-Box Probabilistic Certification","created":"2024-02-17","cited_by_count":4,"area":null},{"id":"2505.11692","title":"The Geometry of ReLU Networks through the ReLU Transition Graph","created":"2025-05-16","cited_by_count":4,"area":null},{"id":"2505.18023","title":"Time to Spike? Understanding the Representational Power of Spiking Neural Networks in Discrete Time","created":"2025-05-23","cited_by_count":4,"area":null},{"id":"2101.05490","title":"Neural networks behave as hash encoders: An empirical study","created":"2021-01-14","cited_by_count":3,"area":null},{"id":"2103.13941","title":"SMILE: Self-Distilled MIxup for Efficient Transfer LEarning","created":"2021-03-25","cited_by_count":3,"area":null},{"id":"2210.05212","title":"On Scrambling Phenomena for Randomly Initialized Recurrent Networks","created":"2022-10-11","cited_by_count":3,"area":null},{"id":"2301.05816","title":"Understanding the Spectral Bias of Coordinate Based MLPs Via Training Dynamics","created":"2023-01-14","cited_by_count":3,"area":null},{"id":"2310.12977","title":"Training Dynamics of Deep Network Linear Regions","created":"2023-10-19","cited_by_count":3,"area":null},{"id":"2310.18725","title":"The Evolution of the Interplay Between Input Distributions and Linear Regions in Networks","created":"2023-10-28","cited_by_count":3,"area":null},{"id":"2404.04662","title":"Learning Minimal Neural Specifications","created":"2024-04-06","cited_by_count":3,"area":null},{"id":"2405.09541","title":"Spectral complexity of deep neural networks","created":"2024-05-15","cited_by_count":3,"area":null},{"id":"2502.04975","title":"Training-free Neural Architecture Search through Variance of Knowledge of Deep Network Weights","created":"2025-02-07","cited_by_count":3,"area":null},{"id":"2502.09954","title":"On Space Folds of ReLU Neural Networks","created":"2025-02-14","cited_by_count":3,"area":null},{"id":"2510.18120","title":"Generalization Below the Edge of Stability: The Role of Data Geometry","created":"2025-10-20","cited_by_count":3,"area":null},{"id":"2601.01417","title":"A Depth Hierarchy for Computing the Maximum in ReLU Networks via Extremal Graph Theory","created":"2026-01-04","cited_by_count":3,"area":null},{"id":"2604.11962","title":"The Linear Centroids Hypothesis: Features as Directions Learned by Local Experts","created":"2026-04-13","cited_by_count":3,"area":null},{"id":"2210.00805","title":"Limitations of neural network training due to numerical instability of backpropagation","created":"2022-10-03","cited_by_count":2,"area":null},{"id":"2307.14823","title":"Fading memory as inductive bias in residual recurrent networks","created":"2023-07-27","cited_by_count":2,"area":null},{"id":"2403.10318","title":"pTNAS: Progressive Neural Architecture Search for Tabular Data","created":"2024-03-15","cited_by_count":2,"area":null},{"id":"2410.16542","title":"A Theoretical Study of Neural Network Expressive Power via Manifold Topology","created":"2024-10-21","cited_by_count":2,"area":null},{"id":"2411.06236","title":"Zero-Shot NAS via the Suppression of Local Entropy Decrease","created":"2024-11-09","cited_by_count":2,"area":null},{"id":"2504.14015","title":"Causal pieces: analysing and improving spiking neural networks piece by piece","created":"2025-04-18","cited_by_count":2,"area":null},{"id":"2505.11370","title":"Understanding Nonlinear Implicit Bias via Region Counts in Input Space","created":"2025-05-16","cited_by_count":2,"area":null},{"id":"2510.12700","title":"Topological Signatures of ReLU Neural Network Activation Patterns","created":"2025-10-14","cited_by_count":2,"area":null},{"id":"2101.07126","title":"A simple geometric proof for the benefit of depth in ReLU networks","created":"2021-01-18","cited_by_count":1,"area":null}],"next_cursor":"100"}