Capacity of the treelike sign perceptrons neural networks with one hidden layer -- RDT based upper bounds
Not on the map. Why?
Abstract (arXiv)
We study the capacity of sign perceptrons neural networks (SPNN) and particularly focus on 1-hidden layer treelike committee machine (TCM) architectures. Similarly to what happens in the case of a single perceptron neuron, it turns out that, in a statistical sense, the capacity of a corresponding multilayered network architecture consisting of multiple sign perceptrons also undergoes the so-called phase transition (PT) phenomenon. This means: (i) for certain range of system parameters (size of data, number of neurons), the network can be properly trained to accurately memorize all elements of the input dataset; and (ii) outside the region such a training does not exist. Clearly, determining the corresponding phase transition curve that separates these regions is an extraordinary task and among the most fundamental questions related to the performance of any network. Utilizing powerful mathematical engine called Random Duality Theory (RDT), we establish a generic framework for determining the upper bounds on the 1-hidden layer TCM SPNN capacity. Moreover, we do so for any given (odd) number of neurons. We further show that the obtained results exactly match the replica symmetry predictions of citeEKTVZ92,BHS92, thereby proving that the statistical physics based results are not only nice estimates but also mathematically rigorous bounds as well. Moreover, for d≤ 5, we obtain the capacity values that improve on the best known rigorous ones of citeMitchDurb89, thereby establishing a first, mathematically rigorous, progress in well over 30 years.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 20 of its 54 references that are in the map. First posted on arXiv 13 Dec 2023.
Builds on
20 of 54 references.
References · 20
- 2016Understanding deep learning requires rethinking generalizationcited by 1,657
- 2018Gradient Descent Provably Optimizes Over-parameterized Neural Networkscited by 655
- 2019Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networkscited by 562
- 2018Learning Overparameterized Neural Networks via Stochastic Gradient Descent on Structured Datacited by 260
- 2019Towards moderate overparameterization: global convergence guarantees for training shallow neural networkscited by 176
- 2018Stochastic Gradient Descent Optimizes Over-parameterized Deep ReLU Networkscited by 139
- 2019Polylogarithmic width suffices for gradient descent to achieve arbitrarily small test error with shallow ReLU networkscited by 120
- 2016Identity Matters in Deep Learningcited by 107
- 2018Small ReLU networks are powerful memorizers: a tight analysis of memorization capacitycited by 76
- 2019Optimization for deep learning: theory and algorithmscited by 70
- 2013Regularly random dualitycited by 23
- 2013Another look at the Gardner problemcited by 23
8 more
- 2019Properties of the geometry of solutions and capacity of multi-layer neural networks with Rectified Linear Units activationscited by 22
- 2023Fully lifted random duality theorycited by 17
- 2013Lifting/lowering Hopfield models ground state energiescited by 13
- 2020Memory capacity of neural networks with threshold and ReLU activationscited by 13
- 2023Binary perceptrons capacity via fully lifted random duality theorycited by 13
- 2013Spherical perceptron as a storage memory with limited errorscited by 11
- 2020Activation function dependence of the storage capacity of treelike neural networkscited by 10
- 2019Mildly Overparametrized Neural Nets can Memorize Training Data Efficientlycited by 7
Built on by
Citing papers per year: 2024: 3.
3 papers.