news
| Jan 22, 2026 | Accepted as a Spotlight at AISTATS 2026 (top 2.5%)! The Good, the Bad, and the Sampled: a No-Regret Approach to Safe Online Classification — joint work with Aldo Pacchiano and Tavor Baharav. |
|---|---|
| Dec 06, 2025 | I’ll present our paper “Safety by Design: High-Probability Constrained Contextual Bandits” with Aldo Pacchiano at the Reliable Machine Learning Workshop at NeurIPS 2025! |
| Nov 06, 2025 | Our joint work Estimating Hitting Times Locally At Scale with Fabian Spaeh, Themis Harris, and Babis Tsourakakis is now accepted at NeurIPS 2025! |
| Oct 01, 2025 | Our joint work with Aldo Pacchiano and Tavor Baharav on “The Good, the Bad, and the Sampled: a No-Regret Approach to Safe Online Classification” is now on arxiv! |