{"total":34,"rows":[{"id":"2307.02129","title":"How Deep Neural Networks Learn Compositional Data: The Random Hierarchy Model","created":"2023-07-05","cited_by_count":71,"area":null},{"id":"2209.09211","title":"Neural Collapse with Normalized Features: A Geometric Analysis over the Riemannian Manifold","created":"2022-09-19","cited_by_count":50,"area":null},{"id":"2310.13620","title":"Bridging Information-Theoretic and Geometric Compression in Language Models","created":"2023-10-20","cited_by_count":26,"area":null},{"id":"2206.12314","title":"Learning sparse features can lead to overfitting in neural networks","created":"2022-06-24","cited_by_count":23,"area":null},{"id":"2305.18512","title":"A Rainbow in Deep Network Black Boxes","created":"2023-05-29","cited_by_count":18,"area":null},{"id":"2410.01444","title":"Geometric Signatures of Compositionality Across a Language Model's Lifetime","created":"2024-10-02","cited_by_count":16,"area":null},{"id":"2310.17715","title":"Outlier Dimensions Encode Task-Specific Knowledge","created":"2023-10-26","cited_by_count":15,"area":null},{"id":"2504.05461","title":"Intermediate Layer Classifiers for OOD generalization","created":"2025-04-07","cited_by_count":14,"area":null},{"id":"2406.01909","title":"A Global Geometric Analysis of Maximal Coding Rate Reduction","created":"2024-06-04","cited_by_count":13,"area":null},{"id":"2105.02468","title":"Relative stability toward diffeomorphisms indicates performance in deep nets","created":"2021-05-06","cited_by_count":12,"area":null},{"id":"2305.19358","title":"Stable Anisotropic Regularization","created":"2023-05-30","cited_by_count":12,"area":null},{"id":"2107.12972","title":"Channel-Wise Early Stopping without a Validation Set via NNK Polytope Interpolation","created":"2021-07-27","cited_by_count":11,"area":null},{"id":"2203.11355","title":"Origami in N dimensions: How feed-forward networks manufacture linear separability","created":"2022-03-21","cited_by_count":9,"area":null},{"id":"2211.13239","title":"Relating Regularization and Generalization through the Intrinsic Dimension of Activations","created":"2022-11-23","cited_by_count":9,"area":null},{"id":"2105.01867","title":"A Theoretical-Empirical Approach to Estimating Sample Complexity of DNNs","created":"2021-05-05","cited_by_count":6,"area":null},{"id":"2303.05161","title":"Inversion dynamics of class manifolds in deep learning reveals tradeoffs underlying generalisation","created":"2023-03-09","cited_by_count":6,"area":null},{"id":"2210.01506","title":"How deep convolutional neural networks lose spatial information with training","created":"2022-10-04","cited_by_count":5,"area":null},{"id":"2303.13635","title":"Low Rank Optimization for Efficient Deep Learning: Making A Balance between Compact Architecture and Fast Training","created":"2023-03-22","cited_by_count":5,"area":null},{"id":"2505.16649","title":"Stochastic Forward-Forward Learning through Representational Dimensionality Compression","created":"2025-05-22","cited_by_count":4,"area":null},{"id":"2110.11400","title":"Channel redundancy and overlap in convolutional neural networks with channel-wise NNK graphs","created":"2021-10-18","cited_by_count":3,"area":null},{"id":"2205.10760","title":"CNNs Avoid Curse of Dimensionality by Learning on Patches","created":"2022-05-22","cited_by_count":3,"area":null},{"id":"2405.06038","title":"From Algorithm to Hardware: A Survey on Efficient and Safe Deployment of Deep Neural Networks","created":"2024-05-09","cited_by_count":2,"area":null},{"id":"2407.17754","title":"DualFed: Enjoying both Generalization and Personalization in Federated Learning via Hierachical Representations","created":"2024-07-25","cited_by_count":2,"area":null},{"id":"2501.02364","title":"Linearly Separable Features in Shallow Nonlinear Networks: Width Scales Polynomially with Intrinsic Data Dimension","created":"2025-01-04","cited_by_count":2,"area":null},{"id":"2605.04358","title":"Intermediate Representations are Strong AI-Generated Image Detectors","created":"2026-05-05","cited_by_count":2,"area":null},{"id":"2107.06409","title":"The Foes of Neural Network's Data Efficiency Among Unnecessary Input Dimensions","created":"2021-07-13","cited_by_count":1,"area":null},{"id":"2309.15328","title":"Exploring Learned Representations of Neural Networks with Principal Component Analysis","created":"2023-09-27","cited_by_count":1,"area":null},{"id":"2311.06816","title":"On original and latent space connectivity in deep neural networks","created":"2023-11-12","cited_by_count":1,"area":null},{"id":"2607.11666","title":"How to Tame Grokking: Representation Geometry as a Control Signal","created":"2026-07-13","cited_by_count":1,"area":null},{"id":"2206.01251","title":"Using Representation Expressiveness and Learnability to Evaluate Self-Supervised Learning Methods","created":"2022-06-02","cited_by_count":0,"area":null},{"id":"2310.01770","title":"A simple connection from loss flatness to compressed neural representations","created":"2023-10-03","cited_by_count":0,"area":null},{"id":"2506.10089","title":"Optimizing Latent Dimension Allocation in Hierarchical VAEs: Balancing Attenuation and Information Retention for OOD Detection","created":"2025-06-11","cited_by_count":0,"area":null},{"id":"2608.10203","title":"A Convolutional Layer Activation Dimensionality Reduction for Out-of-Distribution and Adversarial Attack Detection Methods","created":"2026-08-10","cited_by_count":0,"area":null},{"id":"2609.24688","title":"Understanding Hyperspherical Geometry of ECAPA-TDNN Embedding and Its Impact on Zero-Shot Voice Conversion","created":"2026-09-21","cited_by_count":0,"area":null}],"next_cursor":null}