Tea Time Talks 2024: Alex Lewandowski, Continual Learning, Scalability, and Linearity
Tea Time Talks are back for another year. This summer lecture series, presented by Amii and the RLAI Lab at the University of Alberta, give researchers the chance to discuss early-stage ideas and prospective research. Join us for another series of informal 20-minute talks where AI leaders discuss the future of machine learning research.
Abstract:
In this talk, I outline criteria for a minimal viable algorithm capable of continual learning. Building on the three initial ingredients of continual learning described in Ring's 1994 dissertation, I argue that scalability must be a core component of any such algorithm. I present evidence showing that, unlike nonlinear networks, linear networks meet the criteria for an algorithm capable of continual learning. Going beyond the minimal viable algorithm, I address the representational limitation of linear networks by proposing a new neural network architecture that adaptively balances linearity and nonlinearity.
Смотрите видео Tea Time Talks 2024: Alex Lewandowski, Continual Learning, Scalability, and Linearity онлайн без регистрации, длительностью часов минут секунд в хорошем качестве. Это видео добавил пользователь Amii 27 Сентябрь 2024, не забудьте поделиться им ссылкой с друзьями и знакомыми, на нашем сайте его посмотрели 73 раз и оно понравилось 3 людям.