Skip to content

Open Positions

I have multiple openings for PhD students at UToronto/MPI, Master thesis students at ETH/others, and remote student collaborators. I’m happy to host students who are excited to join the exciting endeavor on LLMs, causality, and multi-agent systems. See my slides for the type of research that I can host projects on.

For applicants who want to impress me, instead of your intro email, you should fill in my forms in detail below, so that I can see many facets of you along dimensions that I really care about! To respect both of our time, I will directly delete any email that shows that a candidate did not read this page carefully. In contrast, if you fill out the form with great detail and do not sound like ChatGPT, you always get a bonus in my impression of you :).

Suitable applicants for all forms have been contacted by 31 Jan, and responses are roughly checked within 2 weeks, except for some very busy periods. Since LLM research changes fast, I suggest you apply no earlier than 1 month before your preferred project starting date.

PhD Student Admission

I am continuously searching for EXCELLENT PhD students who want to work on the research areas highlighted on my main page.

Channel 1: PhD at the University of Toronto

I will look for several PhDs for my Jinesis Lab (pronounced as “Genesis”, in memory of Prof. Patrick Winston) at UofT Computer Science. In the PhD admission at UofT, individual professors have a large decision power.

  • For PhD applicants who have applied in Dec 2024 to {UToronto CS, UToronto Stats Mdoc, ELLIS} programs, I expect the following admission timeline:
    • Feb 19: Verbal offers to clear-cut candidates that I very confidently want to admit. Formal letters will follow
    • Until Apr 15: more interactions and decision-making for waiting-list candidates until final offers. Whenever you have any questions about my admission style, check out “My philosophy for PhD recruiting”, and some other FAQs below on this page.
    • Note: UofT offers are released on a rolling basis: batch 1=Jan 30, batch 2=Feb 19, batch 3=Mar 5, [grad visit days=Mar 17-19], batch 4=Apr 3, last notification batch=Apr 25 – May 15. I will not be in Toronto for the grad visit days, but for admitted students, my fellow profs can host you and you are welcome to video-call me anytime.
  • If you want to apply for a PhD with me in the foreseeable future (i.e., at least one year from now) via the UToronto CS PhD application, it will be an advantage if you have already worked with me. Check out the pre-PhD research form.

Channel 2: PhD via the ELLIS PhD program

As a CIFAR AI Chair, I am listed as an advisor in the ELLIS PhD program too. You can either directly apply to me as your ELLIS main supervisor, or contact me to be a co-supervisor if you are already admitted by a main supervisor. Me and Bernhard Schoelkopf can co-supervise quite several students together, so you are recommended to tag both of us. Check out the ELLIS PhD application by Nov 15.

Channel 3: PhD at Max Planck Institute with Prof Bernhard Schoelkopf, and co-supervised by me

I will co-supervise students together with Prof Bernhard Schoelkopf, so you can directly apply to him and cc’/tag me mentioning that you are interested in co-supervision. Here are some channels to apply, such as the ELLIS PhD program, and IMPRS application to the Max Planck Institute for Intelligent Systems.

Master’s/Undergrad Student Research

Channel 1: ETH students

For ETH Master students, I’m very interested in hosting Master Thesis, Semester Projects, and general research collaboration. Feel free to fill out this form. We can host you under me and one of my co-advisors (either Prof Bernhard Schoelkopf or Prof Mrinmaya Sachan). In special cases, strong ETH undergrad students can apply in the same way too.

Channel 2: Remote Research Mentees

I usually look for students who can devote 6 months to 2 years of time to work with me. You can be of any career stage, from anywhere on the earth. I have past experiences working with master/undergrad/gap-year student, and people who want to do research in their spare time regardless of the day job. Fill out this form.

Students with a diversity background can also try reaching out to PhDs/postdocs of Prof Rada Mihalcea at UMich to see if they have capacities to host projects.

Full-Time Interns

Channel 1: Full-time in-person internship at Max Planck Institute in Tuebingen

Check out the position info here and fill out this form. Bernhard and I can co-supervise you at our EI lab at Max Planck Institute, in Tuebingen, Germany. You can optionally choose to be an intern with us in person, or just remotely collaborate.

Channel 2: Full-time in-person internship at Vector Institute, Toronto, Canada

For people interested in in-person internships from and after Fall 2025, since I am also a faculty member of Vector Institute and CIFAR AI Chair, you can apply to me through the Vector Research Internships.

Postdocs

For postdocs, you can check out these fellowship opportunities to fund your stay at UToronto, and then you can see if our research direction matches via emails to me with the subject “[Postdoc interest match] XXX”.

Tips and FAQs

My philosophy for PhD recruiting

For my upcoming recruiting of PhDs, ideally I hope my lab to have students of diverse expertise: (1) one student experienced with ACL/NeurIPS-type publications, (2) one student very strong in coding, (3) one student with deep philosophical thoughts (with a background in phil/math/physics/etc) and interested in the philosophy of science, and (4) a student with expertise in stats who is either already fluent in or can quickly learn formal causal inference. Above all, I hope that all my students are kind and supportive of each other.

Case studies of successful student applicants:

PhD applicant examples whom I will like very much if they apply to me:

  • Student A is from a country/environment with less NLP resources, but they showed their strong research skills by having 2+ first-author publications at *ACL or NeurIPS. Student A also has a great character to help other people, with experience mentoring more juniors students, and organizing events. Examples: Jiarui Liu and Andrei Muresanu
  • Student B is very good at causal inference. Example: Sean Richardson
  • Student C has extensive past project experiences with researchers in my trusted network, on topics I care about. Their corresponding mentor(s) also sent a very positive strong recommendation message to me. Examples: Amir Zur, and Carrie Yuan
  • Student D is very good at the {history of science, or understanding of political systems across several countries} which can usually reflected by that we can chat for very long without feeling tired. I have a line of research that turns such insights into socially impactful computational projects. Examples: David Guzman (University of Zürich)
  • Student E does not have many research publications, but has a track record to learn anything fast and smartly. Ideally this quality has been demonstrated in our Pre-PhD project together. Example: Samuel Simko
  • Student F shows extraordinary strength in something: e.g., as a software engineer, managing many open-source codebases; a fun NLP+X domain; etc.

Every student is unique. I welcome you to introduce yourself to me as much as possible through the designed questions in my Google form. You could be accepted by unique reasons — congrats! However, if you got declined, then you can always seek inspirations from the above case studies.

Why Pre-PhD research with me before you apply?

The main reason why I suggested pre-PhD research with me is that I can have a better prediction of how life will be with you as my PhD for four or five years. It especially applies to students who are from a diverse set of backgrounds, as part of my fairness pursuit in recruiting.

For PhD admission later, everything is about what reliable and strong signals I can get when predicting the future many years with you. Such signals can be from your papers and recommendation letters when you apply, and can also be from direct interaction with you through a project.

Visa Issues and Collaboration Constraints

Since AI is listed as a sensitive technology, nowadays we have to conform with certain policies and regulations. If you are an individual from Iran, Russia, or China, (1) for remote student collaboration, I am not allowed to collaborate with anyone from the list of sensitive institutions; (2) for any channel that requires a Canadian student visa for you, please be aware of reported cases of student visa delay, especially from sensitive institutions; (2) for any channel that leads to a student visa in Germany, please check ASPI university tracker to see whether your institution will trigger a strong security check for a German visa.

To second-time applicants

If you or applying to me the second time or more, then the most important thing is to highlight your “growth / change” between your recent two applications. For example, you should have an annotated CV contrasting the difference of your old version and the new version (similar to a journal article revision). You can also write a google doc of any skill or experience differences when filling out corresponding answers in the application form. Regardless of whether you get it this time or not, I want to say kudos for your growth!

What if your application does not work out with me?

There are many fantastic opportunities out there as well! If you have general career planning questions, you are very welcome to join our ACL Year-Round Mentorship, for Zoom sessions and slack. Additionally, feel free to check out the research opportunities in the following programs: