FAQ: Applying to my group
FAQ: Applying to my group
I will advertise open positions on the front page of this website. I will keep the size of my group relatively small for the next few years so expect limited positions.
First, you can look at a few recent papers I published. These are usually the best representation of my current research interest. My group primarily works on theoretical aspects of Machine Learning and Reinforcement Learning. For instance, I have no expertise in Deep RL and I do not work with robots in general.
You will need an up-to-date CV highlighting your degrees, your research and professional experience and background. According to the requirements from the doctoral programs I take part in, I will also ask for a motivation letter, ideally in one or two pages maximum, highlighting why you would like to work in my group and 1/ why you want (or feel the need) to have a PhD, 2/ why you think your background and research interest matches mine, and optionally 3/ what specific problems you are interested to work on. There are additional documents such as recommendation letters and track records that need to be uploaded to these platforms.
I work on RL Theory, so a solid background in maths, statistics, statistical learning theory is important to apply to my group. Knowing about RL in general is good, but it could be done though a course on control theory or stochastic processes.
Projects for master thesis or internships can be offered. They will usually be directly related to recent publications from myself or PhD students in the group. As said above, please have a look at my recent publications to get an idea of what are the topics we currently work on. If I advertised a project in a talk, I gave guidelines for how to apply, so please just follow them or check th every bottom of this page.
I am grateful for recommendations from professors who know my work. If you are emailing me on behalf of your professor, please name them and give me some details about your applications. If you have applied to one of the doctoral program I support, you can additionally notify me by email about such recommendation but please do not expect an answer as I cannot reply to all spontaneous applications. Please write an email to applications.vernadelab@gmail.com with the code [ERC25PHD] or [ERC25MSC] if you would like to apply for a Phd or a MSc internship respectively. Please write your motivation in the email and attach a CV and transcript of your grades. If your professor is recommending you, please indicate it clearly in your email (name and contact email of your prof).
The deadline for IMPRS-IS and ELLIS is usually in November. I am aware that this calendar does not match many masters in Europe. At any point in time, if you have made it thus far in this FAQ and you are still motivated to apply to my group, then please write an email to applications.vernadelab@gmail.com with the code [ERC25PHD] or [ERC25MSC] if you would like to apply for a Phd or a MSc internship respectively. Please write your motivation in the email and attach a CV and transcript of your grades. If your professor is recommending you, please indicate it clearly in your email (name and contact email of your prof).
The interviews will be divided into 3 stages, independently of (but in accordance with) the program you've applied to (ELLIS or IMPRS-IS).
Stage 1: Quiz + coding test (30 minutes). Quiz (15-20 minutes): a few technical questions related to the general topics listed in the form. This should be a fast-pace interaction, no long demonstration or discussion needed. Coding (5-10 minute): a short coding test will assess the candidate fluency with basic python / pytorch skills (handling matrices and computing basics stats and visualization)
[Provided the candidate passes the first stage, they will be invited to the second stage]
Stage 2: Paper discussion (45 minute). Each candidate will be sent a short list of papers (3-4), and they will have to choose *one* out of them and read it. The second interview will be a discussion of this paper, including basic understanding questions and more in-depth reasoning and critical discussions about the strengths and weaknesses of the approach within the broader context of the research domain. If the paper is empirical and has code attached, it is expected that the candidate had a look at the code before the interview. If the paper is theoretical and has proofs in Appendix, it is expected that the candidate had a look at the proofs before the interview.
[Provided the candidate passes the second stage, they will be invited to the 3rd stage]
Stage 3: Meeting with the PI. The last step will be a 45 minute discussion with Claire Vernade, about research projects in the lab and the candidate's general research interests. More information about this last interview will be sent to invited candidates.
----
Note 1: IMPRS-IS candidates will also have to give a talk during the Symposium, which the team will attend. We will schedule Stage 2-3 interviews during the symposium to avoid double interviews. If an IMPRS-IS candidate does not pass a Stage-1 interview, they will still participate in the symposium as other PIs may be interviewing them, but they will no longer be considered for my group.