No article text was recorded for this version.
architectural homogeneity constraint (model homogeneity assumption)
Version 1 of 166 papers90 conceptsfrozen 2026-10-04
Papers in this version
- Blog: When Are Bias-Free ReLU Networks Effectively Linear Networks?
- Blog: Learning a Single Neuron with Bias Using Gradient Descent
- Blog: Expressivity of Neural Networks with Fixed Weights and Learned Biases
- Blog: On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models
- Blog: The Convergence Rate of Neural Networks for Learned Functions of Different Frequencies
- Blog: On the Generalization Power of Overfitted Two-Layer Neural Tangent Kernel Models
- Blog: Embedding Principle of Homogeneous Neural Network for Classification Problem
- Blog: Scalar Invariant Networks with Zero Bias
- Blog: Implicit Bias of Gradient Descent for Non-Homogeneous Deep Networks
- Blog: On the Implicit Bias in Deep-Learning Algorithms
- Blog: The dynamics of representation learning in shallow, non-linear autoencoders <sup>*</sup>
- Blog: On the Generalization Power of the Overfitted Three-Layer Neural Tangent Kernel Model
- Blog: Theoretical Investigation of Generalization Bound for Residual Networks
- Blog: Understanding Bias Terms in Neural Representations
- Blog: The Implicit Bias of Adam and Muon on Smooth Homogeneous Neural Networks
- Blog: Sharp bounds for the number of regions of maxout networks and vertices of Minkowski sums
- Blog: A statistical mechanics framework for Bayesian deep neural networks beyond the infinite-width limit
- Blog: The Implicit Bias for Adaptive Optimization Algorithms on Homogeneous Neural Networks
- Blog: An Analytical Formula of Population Gradient for two-layered ReLU network and its Applications in Convergence and Critical Point Analysis
- Blog: Entropy and mutual information in models of deep neural networks
- Blog: Finite Sample Identification of Wide Shallow Neural Networks with Biases
- Blog: Gated Linear Model induced U-net for surrogate modeling and uncertainty quantification
- Blog: Entropy and mutual information in models of deep neural networks
- Blog: Understanding the role of importance weighting for deep learning
- Blog: Fast Axiomatic Attribution for Neural Networks
- Blog: Flavors of Margin: Implicit Bias of Steepest Descent in Homogeneous Neural Networks
- Blog: Convex Geometry of Two-Layer ReLU Networks: Implicit Autoencoding and Interpretable Models
- Blog: Gradient Descent Maximizes the Margin of Homogeneous Neural Networks
- Blog: Penalising the biases in norm regularisation enforces sparsity
- Blog: Deconstructing the Goldilocks Zone of Neural Network Initialization
- Blog: Using Inherent Structures to design Lean 2-layer RBMs
- Blog: The Asymmetric Maximum Margin Bias of Quasi-Homogeneous Neural Networks
- Blog: Analysis of one-hidden-layer neural networks via the resolvent method
- Blog: Gradient Descent Maximizes the Margin of Homogeneous Neural Networks
- Blog: Efficient Algorithms for Learning Depth-2 Neural Networks with General ReLU Activations
- Blog: Generalization Properties of NAS under Activation and Skip Connection Search
- Blog: The late-stage training dynamics of (stochastic) subgradient descent on homogeneous neural networks
- Blog: Agnostic Learning of General ReLU Activation Using Gradient Descent
- Blog: From Complexity to Clarity: Analytical Expressions of Deep Neural Network Weights via Clifford Algebra and Convexity
- Blog: A Random Matrix Perspective on Mixtures of Nonlinearities in High Dimensions
- Blog: Signal Recovery with Non-Expansive Generative Network Priors
- Blog: On feature learning in neural networks with global convergence guarantees
- Blog: Grokking at the Edge of Numerical Stability
- Blog: Low-rank bias, weight decay, and model merging in neural networks
- Blog: Algorithmic Regularization in Learning Deep Homogeneous Models: Layers are Automatically Balanced
- Blog: Implicit Bias of Adversarial Training for Deep Neural Networks
- Blog: Understanding Difficulty-based Sample Weighting with a Universal Difficulty Measure
- Blog: Vector-Valued Variation Spaces and Width Bounds for DNNs: Insights on Weight Decay Regularization
- Blog: Learning Two-layer Neural Networks with Symmetric Inputs
- Blog: The Learning Dynamcis of a Universal Approximator
- Blog: Global Convergence in Training Large-Scale Transformers
- Blog: The Implicit Bias for Adaptive Optimization Algorithms on Homogeneous Neural Networks
- Blog: Training invariances and the low-rank phenomenon: beyond linear networks
- Blog: Representation Costs of Linear Neural Networks: Analysis and Design
- Blog: Large Stepsize Gradient Descent for Non-Homogeneous Two-Layer Networks: Margin Improvement and Fast Optimization
- Blog: Inductive Bias of Gradient Descent for Weight Normalized Smooth Homogeneous Neural Nets
- Blog: Lexicographic and Depth-Sensitive Margins in Homogeneous and Non-Homogeneous Deep Models
- Blog: Tight Bounds on the Smallest Eigenvalue of the Neural Tangent Kernel for Deep ReLU Networks
- Blog: Directional convergence and alignment in deep learning
- Blog: Time/Accuracy Tradeoffs for Learning a ReLU with respect to Gaussian Marginals
Showing 60 of 66 papers.
History & citation
Current · Permanent link
- Incorporating Bias Parameters into Deep Learning Theory merged into this direction (52% overlap)
- Implicit Bias in Non-Homogeneous Neural Networks merged into this direction (57% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_architectural_homogeneity_constraint_mod_1158cb64,
title = {architectural homogeneity constraint (model homogeneity assumption)},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/architectural-homogeneity-constraint-model-homogeneity-assumption/rdr_1158cb642a9a4eb286dc14c3b14fae7a},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}This exact version (version 1, frozen text):
@misc{lacuna_architectural_homogeneity_constraint_mod_1158cb64_v1,
title = {architectural homogeneity constraint (model homogeneity assumption)},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/architectural-homogeneity-constraint-model-homogeneity-assumption/rdr_1158cb642a9a4eb286dc14c3b14fae7a/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}