Welcome!

I am currently an Assistant Professor at the Department of Computer science at Aarhus University, where I am a part of the Algorithms, Data, and Artificial Intelligence section. Previously, I was a postdoctoral fellow at the Institute for Foundations of Machine Learning (UT Austin) and a research fellow at The Simons Institute for the Theory of Computing (UC Berkeley).
I received my PhD from Massachusetts Institute of Technology, advised by Jonathan Kelner and Ronitt Rubinfeld.

Broadly, I work in developing algorithms for machine learning that succeed even when the training data is noisy and unreliable. For example, I have worked on algorithms that are successful even when a small fraction of their training data is altered by a bad actor who is trying to make the algorithm fail.

See this blog post, if you want to know more about some of my recent work.

Feel free to contact me via e-mail: MyFirstNameMyLastName at gmail.com

Publications:

A Fully Polynomial-Time Algorithm for Robustly Learning Halfspaces over the Hypercube, ArXiv
Gautam Chandrasekaran, Adam R. Klivans, Konstantinos Stavropoulos, Arsen Vasilyan
58th ACM Symposium on Theory of Computing (STOC 2026)

Testing Noise Assumptions of Learning Algorithms
Surbhi Goel, Adam R. Klivans, Konstantinos Stavropoulos, Arsen Vasilyan
In submission, arXiv:2501.09189 (2025).
Best Paper Award at Reliable ML from Unreliable Data Workshop @ NeurIPS 2025.

Testable algorithms for approximately counting edges and triangles in sublinear time and space
Talya Eden, Ronitt Rubinfeld, Arsen Vasilyan
17th Innovations in Theoretical Computer Science Conference (ITCS 2026, to appear)

The Power of Iterative Filtering for Supervised Learning with (Heavy) Contamination
Adam R. Klivans, Konstantinos Stavropoulos, Kevin Tian, Arsen Vasilyan
39th Conference on Neural Information Processing Systems (NeurIPS 2025).
Accepted as a spotlight.

Robust learning of halfspaces under log-concave marginals
Jane Lange, Arsen Vasilyan
39th Conference on Neural Information Processing Systems (NeurIPS 2025).
Accepted as a spotlight.

Learning Constant-Depth Circuits in Malicious Noise Models
Adam R. Klivans, Konstantinos Stavropoulos, Arsen Vasilyan
38th Conference on Learning Theory (COLT 2025).

Local Lipschitz Filters for Bounded-Range Functions, ArXiv
Jane Lange, Ephraim Linder, Sofya Raskhodnikova, Arsen Vasilyan
ACM-SIAM Symposium on Discrete Algorithms (SODA 2025).

Tolerant Algorithms for Learning with Arbitrary Covariate Shift
Surbhi Goel, Abhishek Shetty, Konstantinos Stavropoulos, Arsen Vasilyan
38th Conference on Neural Information Processing Systems (NeurIPS 2024).
Accepted as a spotlight.

Efficient Discrepancy Testing for Learning with Distribution Shift
Gautam Chandrasekaran, Adam R. Klivans, Vasilis Kontonis, Konstantinos Stavropoulos, Arsen Vasilyan
38th Conference on Neural Information Processing Systems (NeurIPS 2024).

Plant-and-Steal: Truthful Fair Allocations via Predictions
Ilan Reuven Cohen, Alon Eden, Talya Eden, Arsen Vasilyan
38th Conference on Neural Information Processing Systems (NeurIPS 2024).

Learning Intersections of Halfspaces with Distribution Shift: Improved Algorithms and SQ Lower Bounds
Adam R. Klivans, Konstantinos Stavropoulos, Arsen Vasilyan
37th Conference on Learning Theory (COLT 2024).

Testable Learning with Distribution Shift
Adam R. Klivans, Konstantinos Stavropoulos, Arsen Vasilyan
37th Conference on Learning Theory (COLT 2024).

An Efficient Tester-Learner for Halfspaces
Aravind Gollakota, Adam R. Klivans, Konstantinos Stavropoulos, Arsen Vasilyan
12th International Conference on Learning Representations (ICLR 2024).

Tester-Learners for Halfspaces: Universal Algorithms
Aravind Gollakota, Adam R. Klivans, Konstantinos Stavropoulos, Arsen Vasilyan
37th Conference on Neural Information Processing Systems (NeurIPS 2023).
Accepted for oral presentation.

Agnostic Proper Learning of Monotone Functions: Beyond the Black-box Correction Barrier ArXiv
Jane Lange and Arsen Vasilyan
64th IEEE Symposium on Foundations of Computer Science (FOCS 2023).
Invited to special issue.

Testing Distributional Assumptions of Learning Algorithms, ArXiv
Ronitt Rubinfeld, Arsen Vasilyan
55th ACM Symposium on Theory of Computing (STOC 2023)

Properly Learning Monotone Functions via Local Reconstruction, ArXiv
Jane Lange, Ronitt Rubinfeld, Arsen Vasilyan
63rd IEEE Symposium on Foundations of Computer Science (FOCS 2022)

Monotone Probability Distributions over the Boolean Cube Can Be Learned with Sublinear Samples
Ronitt Rubinfeld, Arsen Vasilyan
11th Innovations in Theoretical Computer Science Conference (ITCS 2020)

Approximating the Noise Sensitivity of a Monotone Boolean Function
Ronitt Rubinfeld, Arsen Vasilyan
Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques
(APPROX/RANDOM 2019).

Doctoral thesis:

Enhancing Learning Algorithms via Sublinear-Time Methods
Arsen Vasilyan, June 2024.