Biography
I am a final year PhD student in the Department of Statistics at the University of Oxford, under the supervision of Patrick Rebeschini and George Deligiannidis. I am funded by EPSRC.
My research interests include reinforcement learning, LLM fine-tuning, optimization and learning theory. In particular, I focus on building and analyzing reinforcement learning algorithms using standard optimization tools, such as natural gradient descent and mirror descent.
Download my CV.
Interests
- Theory of Reinforcement Learning
- Optimization
Education
-
DPhil in Statistics, 2020-present
University of Oxford
-
MSc in Statistical Sciences, 2019-2020
University of Oxford
-
BSc in Statistics, Economics and Finance, 2016-2019
Sapienza University of Rome
Publications
(2025).
Meta-Learning Objectives for Preference Optimization.
arXiv preprint arXiv:2411.06568.
(2025).
Learning mirror maps in policy mirror descent.
To appear in International Conference on Learning Representations.
(2023).
A Novel Framework for Policy Mirror Descent with General Parametrization and Linear Convergence.
Advances in Neural Information Processing Systems.
(2022).
Linear Convergence for Natural Policy Gradient with Log-linear Policy Parametrization.
arXiv preprint: 2209.15382.
(2021).
Dimension-Free Rates for Natural Policy Gradient in Multi-Agent Reinforcement Learning.
arXiv preprint: 2109.11692.
Experience
Applied Scientist Intern
Amazon
I focused on building LLM-based evaluators for the faithfulness of summaries.
Teaching Assistant
University of Oxford
Taught Courses:
- Algorithmic Foundation of Learning
- Advanced Simulation Methods
Supervisor
UNIQ+ DeepMind internship at the University of Oxford
Awards
G-Research Grant for PhD students and postdocs in quantitative fields
EPSRC DTP full scholarship
Full scholarship holder
Honorable mention