Research Area

People excel at real-life tasks that are challenging for modern AI, such as planning and learning new tools from a few examples. How do we solve real-world problems so efficiently?

Our lab builds models of intelligence in humans and machines by developing computational models grounded in AI theory and empirical experimentation. These models provide foundations for training, interpretability, and alignment of AI systems, as well as formal frameworks advancing the science of intelligence.

The lab explores questions such as:
– How do people plan, navigate, and explore in the real world?
– How do people formulate world-models and generate hypotheses about how the world works?
- How can humans and machines work together to invent new concepts and tools?

Working in our lab

Our lab connects a highly collaborative group of researchers at various stages of their careers, and actively engages in national and international collaborations. We use research methods accross artificial intelligence, neuroscience, cognitive science, and psychology.

Lab workflow is often organized around large collaborative projects with weekly online meetings, as the projects connect researchers across countries, institutions and time-zones. The projects are organized around conference dealines, where we may be submitting different iterations of the project to conferences, at various stages of its development. Eventually, a project builds up to a jounal pubication, or to a finalized Ai conference publication, after which it is wrapped up. This is my preferred style of working, as it allows new students quuickly learn research practice. Students are expected to grdully develop independence and led projects of ther own. Some students may working on a project entirely on their own, usually either due to personal preference, or as a temporary project prototyping phase.

Because conference participation in highly important for integrating into the international sceintific community, the lab will prioritize funding travel for major AI/CogSci conferences before other expenses. We may also encorage submitting to non-major conferences, as long as they offer strong networking oportunities (e.g. COSYNE, CCN). Graduate students are encouraged to actively seek various scholarships, fellowships, and stack different sources of funding support -- as is the standard Canadian model. Advisor funding will be offered to lab members, however the Canadian funding structure is not desgned for using a sole source of funing as a living income. Therefore, funding acquisition is a highly important aspect of any Canadian gradiate student, and is treted as a technical skill everyone should learn.

We accept expressions of interest from prospective PhD students and undergraduates interested in joining the lab to build research experience.

While I do not generally offer funding MSc students, I am open to prospective PhD applicants who do not have a MSc, but have prior research experience.

Lab member responsibilities

• Write code, develop and implement computational models
• Prepare weekly (or bi-weekly) slides and progress reports for meetings with academic advisor(s)
• Work effectively with the group toward meetig conference conference deadlines (NeurIPS, ICLR, Cognitive Sciences Society)
• Assist with, and eventually lead, journal paper preparation (i.e., PLOS, Cognition, Nature group's journals)
• Analyze data, fit statistical models, and generate plots
• Contribute to teaching master and bachelor level courses as a teaching assistant
• Design behavioral experiments, and write scripts to process behavioural data
• Present your work as posters and talks, both informally to our collaborators, and at various conferences
• Contribute to a collegial group atmosphere
• Proactively apply for scholarships and seek funding opportunities.

PhD candidate requirements

• Successfully completed a master’s degree (or equivalent research experience) in a quantitaive discipline, such as cognitive psychology, neuroscience, math, physics, computer science or a related disciplines.
• Excellent command of written and spoken English.
• Strong mahematical foundations, an ability or willingness to formulate and prove theorems, and analyze algorithm convergence on the theoretical level. However, if you are a strong psychology or neuroscicnce graduate without a solid math background, we woudl still like to hear from you.
• A record of publications in international-level venues, or pervious research experience with strong references.
• A self-motivated, organized and proactive.
• A working knowledge of programming languages (e.g., Python, R), statistics and AI/ML.
• Familiarity with LLM and scientific computing libraries will be an asset.

Join Us!

Contact us to express interest. Undergraduate applications for directed study and honors projects are welcome at any time.

Please include in your email:

• Which research questions you'd like to work on?
• Which academic papers, books or blog posts inspire you?
• Describe your background in math, psychology, and previous research experience
• Your CV

We do not respond to AI-generated emails, generic inquieries, or emails that do not address the above questions.
We will try to respond to all promising applicants. The easiest way to join this lab as a PhD student is to have worked with us in the past, by contributing to one of our papers. The easiest way to join this lab as an undergraduates is to send us your CV, transcript, and an email outlining your strenth in math, psychology, and prior research experience.