Sledovat
Boris Hanin
Boris Hanin
Assistant Professor, Princeton ORFE
E-mailová adresa ověřena na: princeton.edu - Domovská stránka
Název
Citace
Citace
Rok
Universal function approximation by deep neural nets with bounded width and relu activations
B Hanin
Mathematics 7 (10), 992, 2019
4642019
The principles of deep learning theory
DA Roberts, S Yaida, B Hanin
Cambridge University Press, 2022
3792022
Which neural net architectures give rise to exploding and vanishing gradients?
B Hanin
Advances in neural information processing systems 31, 2018
3562018
Nonlinear Approximation and (Deep) Networks
I Daubechies, R DeVore, S Foucart, B Hanin, G Petrova
Constructive Approximation 55 (1), 127-172, 2022
3392022
How to start training: The effect of initialization and architecture
B Hanin, D Rolnick
Advances in neural information processing systems 31, 2018
3062018
Neural network approximation
R DeVore, B Hanin, G Petrova
Acta Numerica 30, 327-444, 2021
2762021
Complexity of linear regions in deep networks
B Hanin, D Rolnick
International Conference on Machine Learning, 2596-2604, 2019
2722019
Deep relu networks have surprisingly few activation patterns
B Hanin, D Rolnick
Advances in neural information processing systems 32, 2019
2712019
Approximating continuous functions by relu nets of minimal width
B Hanin, M Sellke
arXiv preprint arXiv:1710.11278, 2017
2622017
Finite depth and width corrections to the neural tangent kernel
B Hanin, M Nica
arXiv preprint arXiv:1909.05989, 2019
1842019
Products of many large random matrices and gradients in deep neural networks
B Hanin, M Nica
Communications in Mathematical Physics 376 (1), 287-322, 2020
882020
Scaling limit for the kernel of the spectral projector and remainder estimates in the pointwise Weyl law
Y Canzani, B Hanin
Analysis & PDE 8 (7), 1707-1731, 2015
482015
Random neural networks in the infinite width limit as Gaussian processes
B Hanin
The Annals of Applied Probability 33 (6A), 4798-4819, 2023
472023
Local universality for zeros and critical points of monochromatic random waves
Y Canzani, B Hanin
Communications in Mathematical Physics 378 (3), 1677-1712, 2020
352020
Bayesian interpolation with deep linear networks
B Hanin, A Zlokapa
Proceedings of the National Academy of Sciences 120 (23), e2301345120, 2023
312023
Are more llm calls all you need? towards scaling laws of compound inference systems
L Chen, JQ Davis, B Hanin, P Bailis, I Stoica, M Zaharia, J Zou
arXiv preprint arXiv:2403.02419, 2024
302024
Correlations and pairing between zeros and critical points of Gaussian random polynomials
B Hanin
International Mathematics Research Notices 2015 (2), 381-421, 2015
302015
Pairing of zeros and critical points for random polynomials
B Hanin
282017
Scaling Asymptotics for the Spectral Projector of the Laplacian
Y Canzani, B Hanin
The Journal of Geometric Analysis 28, 111-122, 2018
262018
Correlation functions in random fully connected neural networks at finite width
B Hanin
arXiv preprint arXiv:2204.01058, 10, 2022
252022
Systém momentálně nemůže danou operaci provést. Zkuste to znovu později.
Články 1–20