Vector Research Symposium Highlights Advances in Machine Learning and AI Theory
The Vector Institute held its annual Research Symposium in February, a two-day event dedicated to showcasing the latest developments in machine learning and artificial intelligence. The symposium featured remarks from Vector CEO Garth Gibson and Research Director Richard Zemel, who emphasized the balance between fundamental research and addressing socio-economic issues. Key presentations were delivered by prominent faculty members and CIFAR AI Chairs, including Alan Aspuru-Guzik, Sheila McIlraith, and Nicholas Papernot. A significant theme was the intersection of disciplines, such as the application of robotics in chemistry labs and the integration of privacy guarantees in healthcare AI. Graham Taylor highlighted emerging trends in deep learning theory, specifically the 'double descent' phenomenon, while Anna Golubeva presented research on neural network parameters. Nicholas Papernot introduced Confidential and Private Collaborative Learning (CaPC), a protocol enabling hospitals to collaborate on predictions without sharing sensitive data. The event underscored the importance of theoretical progress for creating trustworthy, reliable, and fair AI systems, demonstrating the interconnected nature of modern AI research across various subfields.
Wire timeline
Vector Research Symposium Highlights Advances in Machine Learning and AI Theory
The Vector Institute held its annual Research Symposium in February, a two-day event dedicated to showcasing the latest developments in machine learning and artificial intelligence. The symposium featured remarks from Vector CEO Garth Gibson and Research Director Richard Zemel, who emphasized the balance between fundamental research and addressing socio-economic issues. Key presentations were delivered by prominent faculty members and CIFAR AI Chairs, including Alan Aspuru-Guzik, Sheila McIlraith, and Nicholas Papernot. A significant theme was the intersection of disciplines, such as the application of robotics in chemistry labs and the integration of privacy guarantees in healthcare AI. Graham Taylor highlighted emerging trends in deep learning theory, specifically the 'double descent' phenomenon, while Anna Golubeva presented research on neural network parameters. Nicholas Papernot introduced Confidential and Private Collaborative Learning (CaPC), a protocol enabling hospitals to collaborate on predictions without sharing sensitive data. The event underscored the importance of theoretical progress for creating trustworthy, reliable, and fair AI systems, demonstrating the interconnected nature of modern AI research across various subfields.
Vector Institute for Artificial Intelligence