Sarath’s note to prospective students:

I am a machine learning generalist and a continual learning specialist. My group explores a broad range of problems in ML, with a particular focus on how learning systems can continually acquire, retain, and build on knowledge.

Here is a list of topics that I am currently interested in:

Continual/lifelong Learning - catastrophic forgetting, loss of plasticity, online learning, unlearning, model merging, memory for continual learning, continual pre-training, continual learning without updating the weights, and knowledge acquisition.
Deep Learning (DL) - New neural network architectures, better recurrent architectures, hybrid architectures, long-term dependency modeling, and memory-augmented architectures.
Optimization - Deep learning dynamics, optimizers for effective exploration of the loss landscape, accelerated training, learning to optimize, optimization for continual learning, optimization for RL, efficient pre-training and fine-tuning algorithms.
Reinforcement Learning (RL) - RL for training LLMs, continual RL, model-based RL, multi-agent RL, world models, world-action models, robotics, RL for drug discovery and materials discovery.
AI for Science - Foundation models for biology (generative and encoder models for proteins, small molecules, DNA, RNA sequences), single-cell foundation models and virtual cells, biological world models, AI/ML/RL for scientific discovery (drug discovery, materials discovery), foundation models for physics/chemistry.
AI Safety - Bias and fairness, trustworthy AI, interpretability, AI alignment, safety in continual learning systems and continual learning for safety.

Fall 2027 Recruitment: I plan to recruit several MSc and PhD students across the research areas listed above. Co-supervision is also possible depending on research fit.

I primarily recruit students through the Mila supervision request program, with a deadline of December 1, 2026. I also recruit PhD students through the ELLIS PhD Fellowship program, with a deadline of October 31, 2026. Students joining through ELLIS will be jointly supervised by me and a Europe-based researcher.

Applicants from all backgrounds are welcome. Women and others from groups underrepresented in machine learning are especially encouraged to apply.

General Expectations:

First and foremost, I am looking for students who are curious, genuinely interested in research, and willing to put in sustained effort.
I look for students with strong foundations in machine learning and deep learning.
You should also have strong programming skills and engineering skills. You should be proficient in Python and autodifferentiation frameworks like PyTorch/JAX.
Some of these technical requirements can be waived if you are an exceptional candidate with a different academic background.

Preparation for the interview:

For the interview, I expect candidates to have strong foundations in the areas listed below. I have included video lectures and textbooks that I consider excellent resources and strongly encourage you to explore. These are broad references, not a reading list to complete: I do not expect you to know everything they cover. The interview will focus on your understanding of fundamental concepts and your ability to reason with them. You are welcome to prepare using other resources.

Linear Algebra - video lectures by Gilbert Strang, essence of linear algebra by 3Blue1Brown.
Probability - Applied Probability video lectures by John Tsitsiklis, Probability Theory video lectures by Krishna Jagannathan.
Optimization - Convex Optimization by Stephen Boyd and Lieven Vandenberghe.
Machine Learning - ML Book - Part 1 and ML Book - Part 2 by Kevin Murphy.
Deep Learning - Understanding Deep Learning by Simon Prince, Dive into Deep Learning by Zhang et al.
Reinforcement Learning - The RL Book by Sutton and Barto (If you are interested in RL).
Large Language Models - Foundations of LLMs by Xiao and Zhu and Language Modeling from Scratch (CS336) course at Stanford (If you are interested in foundation models).

Postdocs

We have three open postdoc positions:

Foundation models for biology
Reinforcement learning, world models, world-action models, and robotics
Foundation models - pre-training, post-training, and safety

Full position descriptions and application instructions are available in the application form. Please apply by November 20, 2026.

Ph.D./Masters Students

The next batch of Ph.D. and Masters admissions will be for Fall 2027.
If you are interested in working with me, you should submit your application to Mila by December 01, 2026. I will only take a look at the applications which mention me as a potential adviser.
I also recruit PhD students through the ELLIS PhD Fellowship program, with a deadline of October 31, 2026. Students joining through ELLIS will be jointly supervised by me and a Europe-based researcher.
If you are shortlisted for an interview, you will need to submit an application to Polytechnique Montréal (with my name). Please note that this application submission does not guarantee admission. Applying to Polytechnique Montréal before the interview helps avoid future admission/visa delays.

Current Students at Polytechnique Montréal and UdeM

If you are already a student (undergrad, Masters, Ph.D.) at Poly or UdeM and want me to be your adviser, email me with your detailed CV, up-to-date copies of all your transcripts, and a summary of your research interests.
If you want me to read your email, your email should have “[RESEARCH_APPLICATION_INTERNAL]” in the subject.

Interns and Visiting Students

Please do not directly e-mail me for visiting positions and internships.
Interested students should complete the following form.
Successful applicants will be contacted directly by one of my students or me.