Matus Telgarsky

Affiliations:

University of Illinois at Urbana-Champaign, Department of Computer Science, IL, USA
Carnegie Mellon University, Machine Learning Department, Pittsburgh, PA, USA

According to our database¹, Matus Telgarsky authored at least 53 papers between 2007 and 2024.

Collaborative distances:

Dijkstra number² of four.
Erdős number³ of three.

Timeline

Legend:

Book

In proceedings

Article

PhD thesis

Dataset

Other

Bibliography

2024

One-layer transformers fail to solve the induction heads task.

[BibT_eX]

[DOI]

Clayton Sanford

Daniel Hsu

Matus Telgarsky

CoRR, 2024

Transformers, parallel computation, and logarithmic depth.

[BibT_eX]

[DOI]

Clayton Sanford

Daniel Hsu

Matus Telgarsky

Proceedings of the Forty-first International Conference on Machine Learning, 2024

Large Stepsize Gradient Descent for Logistic Loss: Non-Monotonicity of the Loss Improves Optimization Efficiency.

[BibT_eX]

[DOI]

Proceedings of the Thirty Seventh Annual Conference on Learning Theory, June 30, 2024

Spectrum Extraction and Clipping for Implicitly Linear Layers.

[BibT_eX]

[DOI]

Ali Ebrahimpour Boroojeny

Matus Telgarsky

Hari Sundaram

Proceedings of the International Conference on Artificial Intelligence and Statistics, 2024

2023

Representational Strengths and Limitations of Transformers.

[BibT_eX]

[DOI]

Clayton Sanford

Daniel J. Hsu

Matus Telgarsky

Proceedings of the Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, 2023

Feature selection and low test error in shallow low-rotation ReLU networks.

[BibT_eX]

[DOI]

Matus Telgarsky

Proceedings of the Eleventh International Conference on Learning Representations, 2023

On Achieving Optimal Adversarial Test Error.

[BibT_eX]

[DOI]

Justin D. Li

Matus Telgarsky

Proceedings of the Eleventh International Conference on Learning Representations, 2023

2022

Feature selection with gradient descent on two-layer networks in low-rotation regimes.

[BibT_eX]

[DOI]

Matus Telgarsky

CoRR, 2022

Convex Analysis at Infinity: An Introduction to Astral Space.

[BibT_eX]

[DOI]

CoRR, 2022

Actor-critic is implicitly biased towards high entropy optimal policies.

[BibT_eX]

[DOI]

Yuzheng Hu

Ziwei Ji

Matus Telgarsky

Proceedings of the Tenth International Conference on Learning Representations, 2022

Stochastic linear optimization never overfits with quadratically-bounded losses on general data.

[BibT_eX]

[DOI]

Matus Telgarsky

Proceedings of the Conference on Learning Theory, 2-5 July 2022, London, UK., 2022

2021

Early-stopped neural networks are consistent.

[BibT_eX]

[DOI]

Ziwei Ji

Justin D. Li

Matus Telgarsky

Proceedings of the Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, 2021

Fast margin maximization via dual acceleration.

[BibT_eX]

[DOI]

Ziwei Ji

Nathan Srebro

Matus Telgarsky

Proceedings of the 38th International Conference on Machine Learning, 2021

Generalization bounds via distillation.

[BibT_eX]

[DOI]

Proceedings of the 9th International Conference on Learning Representations, 2021

Characterizing the implicit bias via a primal-dual analysis.

[BibT_eX]

[DOI]

Ziwei Ji

Matus Telgarsky

Proceedings of the Algorithmic Learning Theory, 2021

2020

Directional convergence and alignment in deep learning.

[BibT_eX]

[DOI]

Ziwei Ji

Matus Telgarsky

Proceedings of the Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, 2020

Neural tangent kernels, transportation mappings, and universal approximation.

[BibT_eX]

[DOI]

Ziwei Ji

Matus Telgarsky

Ruicheng Xian

Proceedings of the 8th International Conference on Learning Representations, 2020

Polylogarithmic width suffices for gradient descent to achieve arbitrarily small test error with shallow ReLU networks.

[BibT_eX]

[DOI]

Ziwei Ji

Matus Telgarsky

Proceedings of the 8th International Conference on Learning Representations, 2020

Gradient descent follows the regularization path for general losses.

[BibT_eX]

[DOI]

Proceedings of the Conference on Learning Theory, 2020

2019

A refined primal-dual analysis of the implicit bias.

[BibT_eX]

[DOI]

Ziwei Ji

Matus Telgarsky

CoRR, 2019

A Gradual, Semi-Discrete Approach to Generative Network Training via Explicit Wasserstein Minimization.

[BibT_eX]

[DOI]

Proceedings of the 36th International Conference on Machine Learning, 2019

Gradient descent aligns the layers of deep linear networks.

[BibT_eX]

[DOI]

Ziwei Ji

Matus Telgarsky

Proceedings of the 7th International Conference on Learning Representations, 2019

The implicit bias of gradient descent on nonseparable data.

[BibT_eX]

[DOI]

Ziwei Ji

Matus Telgarsky

Proceedings of the Conference on Learning Theory, 2019

2018

Risk and parameter convergence of logistic regression.

[BibT_eX]

[DOI]

Ziwei Ji

Matus Telgarsky

CoRR, 2018

Social Welfare and Profit Maximization from Revealed Preferences.

[BibT_eX]

[DOI]

Ziwei Ji

Ruta Mehta

Matus Telgarsky

Proceedings of the Web and Internet Economics - 14th International Conference, 2018

Size-Noise Tradeoffs in Generative Networks.

[BibT_eX]

[DOI]

Bolton Bailey

Matus Telgarsky

Proceedings of the Advances in Neural Information Processing Systems 31: Annual Conference on Neural Information Processing Systems 2018, 2018

2017

Spectrally-normalized margin bounds for neural networks.

[BibT_eX]

[DOI]

Peter L. Bartlett

Dylan J. Foster

Matus Telgarsky

Proceedings of the Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, 2017

Neural Networks and Rational Functions.

[BibT_eX]

[DOI]

Matus Telgarsky

Proceedings of the 34th International Conference on Machine Learning, 2017

Non-convex learning via Stochastic Gradient Langevin Dynamics: a nonasymptotic analysis.

[BibT_eX]

[DOI]

Maxim Raginsky

Alexander Rakhlin

Matus Telgarsky

Proceedings of the 30th Conference on Learning Theory, 2017

2016

Greedy bi-criteria approximations for k-medians and k-means.

[BibT_eX]

[DOI]

Daniel J. Hsu

Matus Telgarsky

CoRR, 2016

Rate of Price Discovery in Iterative Combinatorial Auctions.

[BibT_eX]

[DOI]

Jacob D. Abernethy

Sébastien Lahaie

Matus Telgarsky

Proceedings of the 2016 ACM Conference on Economics and Computation, 2016

benefits of depth in neural networks.

[BibT_eX]

[DOI]

Matus Telgarsky

Proceedings of the 29th Conference on Learning Theory, 2016

2015

Convex Risk Minimization and Conditional Probability Estimation.

[BibT_eX]

[DOI]

Matus Telgarsky

Miroslav Dudík

Robert E. Schapire

CoRR, 2015

Representation Benefits of Deep Feedforward Networks.

[BibT_eX]

[DOI]

Matus Telgarsky

CoRR, 2015

Convex Risk Minimization and Conditional Probability Estimation.

[BibT_eX]

[DOI]

Matus Telgarsky

Miroslav Dudík

Proceedings of The 28th Conference on Learning Theory, 2015

Tensor Decompositions for Learning Latent Variable Models (A Survey for ALT).

[BibT_eX]

[DOI]

Proceedings of the Algorithmic Learning Theory - 26th International Conference, 2015

2014

Tensor decompositions for learning latent variable models.

[BibT_eX]

[DOI]

Animashree Anandkumar

J. Mach. Learn. Res., 2014

Scalable Nonlinear Learning with Adaptive Polynomial Expansions.

[BibT_eX]

[DOI]

CoRR, 2014

Scalable Non-linear Learning with Adaptive Polynomial Expansions.

[BibT_eX]

[DOI]

Proceedings of the Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014, 2014

2013

Duality and Data Dependence in Boosting /.

[BibT_eX]

[DOI]

Matus Telgarsky

PhD thesis, 2013

Dirichlet draws are sparse with high probability

[BibT_eX]

[DOI]

Matus Telgarsky

CoRR, 2013

Moment-based Uniform Deviation Bounds for $k$-means and Friends.

[BibT_eX]

[DOI]

Matus Telgarsky

Sanjoy Dasgupta

CoRR, 2013

Moment-based Uniform Deviation Bounds for <i>k</i>-means and Friends.

[BibT_eX]

[DOI]

Matus Telgarsky

Sanjoy Dasgupta

Proceedings of the Advances in Neural Information Processing Systems 26: 27th Annual Conference on Neural Information Processing Systems 2013. Proceedings of a meeting held December 5-8, 2013

Margins, Shrinkage, and Boosting.

[BibT_eX]

[DOI]

Matus Telgarsky

Proceedings of the 30th International Conference on Machine Learning, 2013

Boosting with the Logistic Loss is Consistent.

[BibT_eX]

[DOI]

Matus Telgarsky

Proceedings of the COLT 2013, 2013

2012

A Primal-Dual Convergence Analysis of Boosting.

[BibT_eX]

[DOI]

Matus Telgarsky

J. Mach. Learn. Res., 2012

Statistical Consistency of Finite-dimensional Unregularized Linear Classification

[BibT_eX]

[DOI]

Matus Telgarsky

CoRR, 2012

Agglomerative Bregman Clustering.

[BibT_eX]

[DOI]

Matus Telgarsky

Sanjoy Dasgupta

Proceedings of the 29th International Conference on Machine Learning, 2012

2011

Blackwell Approachability and Minimax Theory

[BibT_eX]

[DOI]

Matus Telgarsky

CoRR, 2011

The Convergence Rate of AdaBoost and Friends

[BibT_eX]

[DOI]

Matus Telgarsky

CoRR, 2011

The Fast Convergence of Boosting.

[BibT_eX]

[DOI]

Matus Telgarsky

Proceedings of the Advances in Neural Information Processing Systems 24: 25th Annual Conference on Neural Information Processing Systems 2011. Proceedings of a meeting held 12-14 December 2011, 2011

2010

Hartigan's Method: k-means Clustering without Voronoi.

[BibT_eX]

[DOI]

Matus Telgarsky

Andrea Vattani

Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, 2010

2007

Signal Decomposition using Multiscale Admixture Models.

[BibT_eX]

[DOI]

Matus Telgarsky