Skip to main content
Grigoris Chrysos

Grigorios Chrysos

Assistant Professor

My research focuses on reliable machine learning and the design and study of expressive models that are robust to noise and generalize well in out-of-distribution data. Concretely:

I am interested in understanding the inducative bias of deep networks and properties of existing architectures through empirical and theoretical studies.

I am interested in the complete theoretical understanding of (neural/polynomial) networks, including their expressivity, trainability, generalization properties.

The understanding of the inductive bias will enable us to design improved networks. Towards that end, I have worked extensively on polynomial networks (PNs). PNs that capture high-degree interactions between inputs.

Machine learningI am interested in the extrapolation properties of existing networks and improving their performance, especially in the context of conditional generative models. In the short-term, I will continue to explore the robustness of these models to malicious attacks, as well as the impact of adversarial perturbations on different classes. In the long-term, I plan to design models that are both robust and fair, and can generalize well to unseen combinations.

Department

Electrical & Computer Engineering

Contact

M1002H, Engineering Centers Building
1550 Engineering Dr
Madison, WI

  • PhD 2020, Imperial College London
  • M. Eng. 2014, National Technical University of Athens (NTUA)

  • Machine learning
  • Deep learning

  • 2024 Gemma, Access to Gemma Academic Program
  • 2024 OpenAI, Access to OpenAI Researcher Access Program
  • 2023 DAAD, DAAD AInet Fellow (prestigious German program)
  • 2022 Neurl, Best reviewer award
  • 2022 ICLR, Best reviewer award
  • 2021 IMCL, Best reviewer award
  • 2019 Amazon Cloud and Nvidia GPU, Amazon Cloud credits and Nvidia GPU donations for conducting research
  • 2019 ICLR, Travel Award
  • 2019 MDPI journal, Travel Award
  • 2017 Imprerial College London, Winner of the annual innovation competition
  • 2015 Engineering & Physical Sciences Research Council of United Kingdom (UK), 4-year Scholarship
  • 2008 University and state scholarships

  • Panagakis, Y., Kossaifi, J., Chrysos, G., Oldfield, J., Patti, T., Nicolaou, M. A., Anandkumar, A., & Zafeiriou, S. (2024). Tensor methods in deep learning. In Signal Processing and Machine Learning Theory (pp. 1009–1048). Elsevier.
  • Rocamora, E. A., Liu, F., Chrysos, G., Olmos, P. M., & Cevher, V. (2024). Efficient local linearity regularization to overcome catastrophic overfitting. In International Conference on Learning Representations (ICLR).
  • Chen, Y., Liu, F., Lu, Y., Chrysos, G., & Cevher, V. (2024). Generalization of scaled deep ResNets in the mean-field regime. In International Conference on Learning Representations (ICLR).
  • Deschenaux, J., Krawczuk, I., Chrysos, G., & Cevher, V. (2024). Going beyond compositional generalization, DDPMs can produce zero-shot interpolation. In International Conference on Machine Learning (ICML).
  • Puigdemont, P., Skoulakis, S., Chrysos, G., & Cevher, V. (2024). Learning to Remove Cuts in Integer Linear Programming. In International Conference on Machine Learning (ICML).
  • Chatziagapi, A., Chrysos, G., & Samaras, D. (2024). MIGS: Multi-Identity Gaussian Splatting via Tensor Decomposition. In European Conference on Computer Vision (ECCV).
  • Oldfield, J., Georgopoulos, M., Chrysos, G., Tzelepis, C., Panagakis, Y., Nicolaou, M. A., Deng, J., & Patras, I. (2024). Multilinear Mixture of Experts: Scalable Expert Specialization through Factorization. In Neural Information Processing Systems (NeurIPS).
  • Chen, Y., Chrysos, G., Georgopoulos, M., & Cevher, V. (2024). Multilinear Operator Networks. In International Conference on Learning Representations (ICLR).
  • Afzal, A., Chrysos, G., Cevher, V., & Shoaran, M. (2024). REST: Efficient and Accelerated EEG Seizure Analysis through Residual State Updates. In International Conference on Machine Learning (ICML).
  • Abad Rocamora, E., Wu, Y., Liu, F., Chrysos, G., & Cevher, V. (2024). Revisiting Character-level Adversarial Attacks for Language Models. In International Conference on Machine Learning (ICML).

  • COMP SCI 761 - Mathematical Foundations of Machine Learning (Spring 2025)
  • E C E 399 - Independent Study (Spring 2025)
  • E C E 699 - Advanced Independent Study (Spring 2025)
  • E C E 761 - Mathematical Foundations of Machine Learning (Spring 2025)
  • E C E 790 - Master's Research (Spring 2025)
  • E C E 999 - Advanced Independent Study (Spring 2025)
  • COMP SCI 532 - Matrix Methods in Machine Learning (Fall 2024)
  • E C E 532 - Matrix Methods in Machine Learning (Fall 2024)
  • E C E 699 - Advanced Independent Study (Fall 2024)
  • E C E 790 - Master's Research (Fall 2024)
  • M E 532 - Matrix Methods in Machine Learning (Fall 2024)
  • E C E 204 - Data Science & Engineering (Spring 2024)
  • E C E 890 - Pre-Dissertator's Research (Spring 2024)