Busra Tugce Gurbuz

Ph.D. Candidate ⎮ AI Researcher ⎮ Neuroscience-Inspired AI ⎮ Builder of Human-Centric Intelligence

Hi, I’m Tuğçe [ˈtuːtʃe] ! 👋

I’m currently pursuing my Ph.D. at Mila-Quebec AI Institute, McGill University, and Montreal Neurological Institute — institutions that continuously challenge and inspire me to bridge the gap between brains and machines.

🧠 My Research

My work integrates neuroscience, psychology, and artificial intelligence, with a focus on reinforcement learning (RL) and representation learning. I’m particularly interested in developing compact, structured, and agent-centric world models that support sample-efficient learning, generalization, and planning.

As RL expands into high-stakes domains like robotics, healthcare, and embodied agents, such representations become critical for building learning systems that are both tractable and trustworthy. Inspired by how humans form internal models of the world, I explore ways to integrate contrastive learning and multimodal large-scale models into scalable RL systems that better reflect how intelligent behavior emerges.

My research receives generous support from Fonds de recherche du Québec and Unifying Neuroscience and Artificial Intelligence - Québec scholarships.

🌍 Responsible & Impactful AI

Beyond technical innovation, I’m driven by the question: “How can we ensure AI truly benefits society?” I’m deeply invested in AI safety, ethical development, and socially impactful applications, particularly in domains like healthcare, mental health, and human-AI collaboration. If you’re working on projects aligned with these goals, I’d love to connect — feel free to reach out.

👩🏻‍🎓 Before starting my Ph.D. program

I was an undergraduate student in the Psychology Department at Bilkent University(Ankara, Turkey) with a computational and cognitive neuroscience specialization. I also conducted research on computational neuroscience and psychology at Aysel Sabuncu Brain Research Center (UMRAM) where I used artificial neural networks to model one of the visual illusions we experience. 🔮

🌟 My Approach

Grounded in formal scientific training through my BSc in neuroscience/psychology and PhD in AI degrees, I bring a strong emphasis on precision, methodological rigor, and long-term research impact. These principles, combined with the iterative and exploratory mindset of machine learning, shape how I frame questions, build models, and collaborate on research.

🌸 Beyond the Research

When I’m not thinking about how brains and machines learn, you’ll likely find me:

• Cooking dishes from around the world — 40 countries and counting!
• Playing bass guitar (once a professional musician, now a home-musician).
• Strength training and experimenting with calisthenics and vertical pole fitness.
• Traveling, exploring new cultures, and seeking inspiration from unexpected places.

📑 Blog Posts

After years of dreaming, I’ve finally found the courage to launch my own blog. I’m excited to introduce you to the Robot Psychologist’s Blog!

📖 Check here for recent posts

🗄 Not-Published Projects

Some class/side projects that almost made it to a conference submission if only we had infinite compute, zero schedule conflicts, and maybe a few more hours in the day 🥹

✨✨ Not every idea has to become a paper (or survive Reviewer #2), sometimes, the best part is just diving in and seeing where curiosity takes you ✨✨

But hey, I still have hope! (That’s why I put them before the actual published work section below 😉) If any of these ideas catch your eye and you want to pick them up, let me know, I’m always up for a collaboration (or just chatting about wild project ideas).

- Vision-Language Models & Reasoning

• Improving Compositional Reasoning of Vision Language Models via Multi-Agent Collaboration
  B.T. Gurbuz, B.L. Dang, Q. Yang
  Links between Computer Vision and Language @Mila,Winter-2025
  Blog Post is coming soon...
  

  

- Offline Reinforcement Learning

• Hiearchical Memory Decision Transformers
  B.T. Gurbuz
  Reinforcement Learning @Mila,Winter-2024
  Blog Post is coming soon...
  

  

📰 Published Projects

- Reinforcement Learning

• Foraging minds: Exploring memory mechanisms for continual reinforcement learning
  B.T. Gurbuz, A. Pruss, N.Anand, S. Krishna , E.B. Muller, C. Pack
  Proceedings of Pucón Learning and AI Summit (PLENA) 2024
  

  

- AI Agents & Safety

• SandboxSocial: A Sandbox for Social Media Using Multimodal AI Agents
  M.P. Touzel*, S. Sarangi*, G. Krishnakumar*, B.T. Gurbuz, A. Welch, Z. Yang, A. Musulan, H. Yu, E. Kosak-Hine, T. Gibbs, C. Thibault, R. Rabbany, JF. Godbout, D. Zhao, K. Pelrine.
  International Joint Conference on Artificial Intelligence (IJCAI) 2025 & The 63rd Annual Meeting of the Association for Computational Linguistics (ACL) NLP4PI 2025 Workshop
  

  

• Simulation System Towards Solving Societal-Scale Manipulation
  M.P. Touzel*, S. Sarangi*, A. Welch*, G. Krishnakumar, D. Zhao Z. Yang, A. Musulan, H. Yu, E. Kosak-Hine, T. Gibbs, C. Thibault, B.T. Gurbuz, R. Rabbany, JF. Godbout, K. Pelrine.
  Neural Information Processing Systems (NeurIPS) 2024 Safe GenAI Workshop
  

  

- AI for Neuroscience

• Learning to combine top-down context and feed-forward representations under ambiguity with apical and basal dendrites
  N. Islah, G. Etter, M. Tugsbayar, B.T. Gurbuz, B. Richards , E.B. Muller
  Cerebral Cortex, Volume 35, Issue 6, June 2025
  

  

- Computational Neuroscience

• Abrupt Learning: How Does the Brain Decide What and When to Learn?
  B.T. Gurbuz *, A. Pruss*, E.B. Muller, S. Krishna, C. Pack
  Proceedings of European Conference on Visual Perception (ECVP) 2024
  

  

• Tilt aftereffect spreads across the visual field
  B.T. Gurbuz, H. Boyaci
  Vision Research, 205, 108174, 2023