Q&A with Vector AI scholarship winners

Friday, June 14, 2024

Three incoming computer science master’s students, Benjamin Schneider, Hala Sheta and Xin Yan, have been named recipients of this year’s Vector Institute Scholarships in Artificial Intelligence (VSAI).

Launched in 2018, the VSAI program supports top students enrolled in Vector-recognized master’s programs or pursuing individualized artificial intelligence (AI) study paths at universities across Ontario. The 2024-25 cohort was valued at around $2 million, with each student receiving $17,500. Beyond financial aid, they can access career opportunities such as Vector’s Digital Talent Hub, networking events, research talks, or professional development programs.

banner photo of scholarship winners. In between each people's faces theres a white border/frame

Benjamin Schneider (left), Hala Sheta (middle), and Xin Yan (right) will conduct cutting-edge research in AI at the Cheriton School of Computer Science. 

We connected with the students to discuss their upcoming research plans and thoughts on receiving the scholarship.

Benjamin Schneider

Where did you complete your undergraduate studies?

I completed my undergraduate in computer science at the University of Manitoba, from 2018 to 2023.

What is your main research focus?

My research focuses on controlling how foundational models use knowledge, especially how external information sources can be integrated into models while maintaining control over how these sources are used. I am also interested in methods for measuring the reliability and robustness of foundational models. I am co-supervised by Professors Florian Kerschbaum and Wenhu Chen.

What real-life issues is your research trying to tackle?

My research centres around making foundational models easier to update and deploy. Integrating new information into a model, especially sensitive information, is computationally costly and can easily leak private information. My work seeks to make such updates efficient and provide tools to control how the new knowledge is used by the model. Having better tools for integrating knowledge into models allows for more information sources to be used while ensuring responsible data handling.

What made you choose UWaterloo for grad studies?

Under UWaterloo’s Undergraduate Research Fellowship (URF) program, I spent a summer working in Waterloo as a researcher. This experience convinced me that I enjoy research and that I should pursue grad school. So, when it came time to choose a graduate program, UWaterloo was the obvious choice!

Hala Sheta

Where did you complete your undergraduate studies?

From 2019 to 2023, I attended the University of Toronto. I completed a Bachelor of Science in Computational Cognitive Science, with minors in Computer Science and Linguistics.  

What is your main research focus?

My research focuses on the computational modelling of cognition, specifically through the medium of natural language. I will work with Professors Freda Shi and Dan Brown to study how language is employed and used in artificial machines, its implications on a machine’s semantic understanding and “cognitive ability,” and possible applications in ill-defined domains such as Computational Creativity.

What real-life issues is your research trying to tackle?

This research attempts to tackle the imminent black-box problem that is prevalent in the AI field, where the focus is mostly on the output rather than the computational process. By unpacking and focusing on the ins and outs of the computational process, we can reach a model that cognizes in a “human-like” manner and by proxy, discover something about the machinery fueling our own cognition.

What made you choose UWaterloo for grad studies?

I chose UWaterloo for my graduate studies because of its cutting-edge research facilities and its network of top experts in my field. The university’s rich environment of multi-disciplinary research and technology aligns with my academic goals, and I want to contribute to that circle of knowledge and research.

How do you feel about being a recipient?

I feel immensely grateful that I am able to receive support in pursuing my academic goals. This opportunity motivates me to aim higher in my academic pursuits, and I am committed to making the most out of it.

Xin Yan

Where did you complete your undergraduate studies?

I am from China, and I studied at Wuhan University’s School of Computer Science from 2019 to 2023. I obtained a Bachelor of Engineering in Computer Science and Technology.

What is your main research focus?

My past research has focused on evaluating and improving reasoning across different modalities, including vision, language, audio, and motion. I have designed multimodal benchmarks for evaluating visual reasoning abilities, generated symbolic programs for hierarchical reasoning and planning, and developed an image modeling algorithm that draws on the relationships between language tokens. Additionally, I have introduced a method to unify multiple modalities and tasks in a single model.

In my graduate research, I plan to improve the reasoning abilities of multimodal AI models, especially in generating new knowledge from their observations. I will focus on this topic with Professor Yuntian Deng in my two-year master's study at the University of Waterloo.

What real-life issues is your research trying to tackle?

My research is related to Large Language Models (LLM), which are widely used in everyone's daily life, even though they may not know AI. LLMs are powerful and can reach or even surpass human beings in many tasks. However, all these models are still trying to fit the distribution of human data, and can hardly generate new knowledge. I hope to give machines human-level reasoning abilities, so they can quickly adapt to changing environments and innovate in science and technology. With stronger reasoning abilities, we could develop a human-level chatbot, which can learn new knowledge about the world by interacting with users and make scientific discoveries like a human scientist.

What made you choose UWaterloo for grad studies?

Firstly, UWaterloo is a globally recognized institution, renowned particularly for its excellence in Computer Science. Its reputation for innovation and research is well-deserved and widely acknowledged around the world. Secondly, the MMath program at UWaterloo particularly appeals to me due to its high-quality curriculum and rich research environment. Also, the program offers some financial support such as the Graduate Teaching Assistantships and Graduate Research Assistantships, which helps me focus on my studies and research. Finally, the opportunity to collaborate with and learn from Professor Yuntian Deng was a major draw for me. His research is fascinating and aligns closely with my research interests. I believe I can achieve significant research breakthroughs under his guidance. Therefore, I am excited to be part of the vibrant community at UWaterloo.

How do you feel about being a recipient?

I feel incredibly honoured and grateful to be a recipient. It's a testament to the hard work and dedication I've put into my studies. The Vector Scholarship will support my graduate study and research in many ways, not only for the financial support that enables me to fully concentrate on my research without financial concerns but also for other opportunities provided by the Vector Institute. I believe I can contribute more with this support.

  1. 2024 (68)
    1. July (11)
    2. June (11)
    3. May (15)
    4. April (9)
    5. March (13)
    6. February (1)
    7. January (8)
  2. 2023 (70)
    1. December (6)
    2. November (7)
    3. October (7)
    4. September (2)
    5. August (3)
    6. July (7)
    7. June (8)
    8. May (9)
    9. April (6)
    10. March (7)
    11. February (4)
    12. January (4)
  3. 2022 (63)
    1. December (2)
    2. November (7)
    3. October (6)
    4. September (6)
    5. August (1)
    6. July (3)
    7. June (7)
    8. May (8)
    9. April (7)
    10. March (6)
    11. February (6)
    12. January (4)
  4. 2021 (64)
  5. 2020 (73)
  6. 2019 (90)
  7. 2018 (82)
  8. 2017 (51)
  9. 2016 (27)
  10. 2015 (41)
  11. 2014 (32)
  12. 2013 (46)
  13. 2012 (17)
  14. 2011 (20)