user_image

Matteo Boglioni

Computer Science MSc Student at ETH Zürich.
I’m particularly interested in red-teaming and model auditing as proactive techniques to uncover existing models’ vulnerabilities. I aim to assess and expose these flaws and inconsistencies, both in performance and security, to increase robustness and reliability towards distribution shifts, adversarial inputs, and unintended behaviors. I believe this step is crucial for a responsible and ethical deployment of AI.

📄 Resume 🔍 Research Statement

EXPERIENCE

  • 2025-03 – 2025-09

    Carnegie Mellon University, Visiting Research Scholar

  • 2023 – 2025

    ETH Zürich, Computer Science MSc, Machine Intelligence

  • 2023-01 – 2023-05

    University Of Wisconsin-Madison, ECE Visiting Student

  • 2020 – 2023

    Politecnico di Milano, Engineering Of Computing Systems BSc

RESEARCH INTERESTS

  • Machine Learning Privacy and Security
  • LLMs Robustness and Generalization
  • Privacy

HOBBIES

  • CLIMBING

    85%

  • ROAD CYCLING

    75%

  • VIDEOMAKING

    60%

  • TRAVELING & BACKPACKING

    100%
-->

RECENT PUBLICATIONS

[All Publications]
  1. Enhancing One-run Privacy Auditing with Quantile Regression-Based Membership Inference
    In Workshop on Theory and Practice of Differential Privacy, 2025
  2. Optimizing Canaries for Privacy Auditing with Metagradient Descent
    Boglioni, Matteo, Liu, Terrance, Ilyas, Andrew, and Wu, Zhiwei Steven
    2025