Lorenzo Pacchiardi

Department of Statistics

University of Oxford

24-29 St Giles'

Oxford OX1 3LB

United Kingdom

I work on the safety of large language models with Dr Owain Evans and consult on standards for high-risk AI systems for the European AI Act at the Future of Life Institute. I am broadly interested in technical and policy ways to address the societal impact of AI.

During my PhD in Statistics and Machine Learning at Oxford, I mostly worked on inference for generative models, spanning Bayesian simulation-based inference and training methods for generative neural networks. Some of the approaches I developed build on concepts from probabilistic forecasting. My supervisors were Prof. Ritabrata Dutta (Uni. Warwick) and Prof. Geoff Nicholls (Uni. Oxford). I was also interested in conformal prediction, normalizing flows, gradient-based MCMC methods and weather forecasting.

Prior to my PhD studies, I obtained a Bachelor’s degree in Physical Engineering from Politecnico di Torino (Italy). Afterwards, I studied towards an MSc in Physics of Complex Systems awarded by Politecnico di Torino and Université Paris-Sud, France. I have carried out my MSc thesis at LightOn, a machine learning startup in Paris.


Dec 20, 2022 I won a travel award at the International Conference for Statistics and Data Science by the Institute of Mathematical Statistics. I really enjoyed the opportunity to present my work there!
Nov 17, 2022 On the 9th of November, I successfully defended my PhD thesis :champagne:. A profound thanks to my examiners Chris Holmes and Chris Oates for their insightful comments and suggestions.
Jun 3, 2022 I have two preprints on arXiv exploring an alternative to adversarial training to train generative networks; the first applies it to probabilistic forecasting, while the second is concerned with Likelihood-Free Inference. I feel that is a promising overlooked approach. Happy to hear any feedback! :smile:
Mar 16, 2022 I have created a website aimed at listing works on using Neural Networks in Bayesian Likelihood-Free Inference :nerd_face: I have added the papers I am aware of, but please contribute others you may know (see “About” page for how).
Feb 3, 2022 Our paper discussing a new method for learning summary statistics for Approximate Bayesian Computation has been published in Journal of Machine Learning Research! :champagne:

selected publications and preprints

  1. arXiv
    Likelihood-Free Inference with Generative Neural Networks via Scoring Rule Minimization
    Pacchiardi, Lorenzo, and Dutta, Ritabrata
    arXiv preprint arXiv:2205.15784 2022
  2. arXiv
    Probabilistic Forecasting with Generative Networks via Scoring Rule Minimization
    Pacchiardi, Lorenzo, Adewoyin, Rilwan, Dueben, Peter, and Dutta, Ritabrata
    arXiv preprint arXiv:2112.08217 2022
  3. JMLR
    Score Matched Neural Exponential Families for Likelihood-Free Inference
    Pacchiardi, Lorenzo, and Dutta, Ritabrata
    Journal of Machine Learning Research 2022
  4. JSS
    ABCpy: A High-Performance Computing Perspective to Approximate Bayesian Computation
    Dutta, Ritabrata, Schoengens, Marcel, Pacchiardi, Lorenzo, Ummadisingu, Avinash, Widmer, Nicole, Künzli, Pierre, Onnela, Jukka-Pekka, and Mira, Antonietta
    Journal of Statistical Software 2021