Talks and presentations

Guest lecture on the deep learning toolkit covering common frameworks and practical tooling for building and training deep learning models.

Presented our work “Bayesian Adaptation of Network Depth and Width for Continual Learning”

Talked about recent progress in test-time adaptation and continual test-time adaptation.

Talked about the technical details of OpenAI DALL-E, a text-to-image generation model able to combine unrelated concepts in plausible ways, anthropomorphize objects and animals, and render text in images.

Gave an introduction to Natural Language Processing (NLP), related problems, data representation for deep learning, and common deep-learning models in the domain.