Thibaut Lienart

I'm a fourth year PhD student in computational statistics at the University of Oxford, now based at the Alan Turing Institute for my last year. You can find some informations about myself and my research interests here. Alternatively, you can find me on LinkedIn (here).
I also (used to) keep a small blog with some research notes which I'm hoping might be useful for other students someday.

New twitter account (I'm still working it out): @t_lienart.


  • (Spring 2017) I am based at the ATI and working on implementing a toolbox in Julia for rejection free samplers such as the Bouncy Particle Sampler and the Zig Zag Sampler.

  • (Spring 2017) check out our new group website: OxCSML (Computational Statistics, Machine Learning and Statistical Methodology).

  • (February 2017) our paper (on Distributed Bayesian Inference with EP) has been accepted for publication in the Journal of Machine Learning Research.

  • (December 2016) I will be going to NIPS 2016, our last paper on distributed Bayesian inference got accepted for two workshops (AABI2016 (Advances in Approximate Bayesian Inference) and BDL2016 (Bayesian Deep Learning))

  • (November 2016) joined the Alan Turing Institute in London as a senior PhD student.

  • (September 2016) second version of our paper suggesting a new method for training large-scale models with a distributed architecture, check it out! link

  • (June 2016) invited speaker at the Swiss statistical office, perspectives on big data and machine learning research.

  • (Spring 2016) I'll be an intern at Amazon Research Berlin over the summer.

  • (Spring 2016) confirmed my involvement with the Cambridge Coding Academy startup (now Cambridge Spark).

  • (January 2016) after about 6 months of work, we put a pre-print on ArXiv suggesting a new method for training large-scale models (e.g., Deep Neural Networks) with a distributed architecture, check it out! (cf. updated link above)

  • (Winter 2015) while in the US prior to Montreal, I briefly visited Barbara Engelhardt's group at Princeton and David Blei's group at Columbia. Great people and exciting research!

  • (Fall 2015) I helped in the organization of the Scalable Monte Carlo workshop held at NIPS’15. We were lucky to have excellent speakers and very interesting posters!

  • (Summer 2015) we submitted a paper to NIPS’15 on how to improve Particle Belief Propagation using Expectation Propagation: corrected version here, code here. Update: submission accepted, I presented it at the SMC2015 conference in Paris and at NIPS’15 (link to poster). See also corresponding blog post.

  • (Spring 2015) with Adrien Hitz we organized a series of graduate lectures on Graphical Models. More informations here.

Contact details

Thibaut lienart 

Thibaut Lienart (,
Department of Statistics,
University College, University of Oxford.

PhD. supervisors: