Imperial College ICARL Seminars - Jake Bruce (DeepMind)

We are excited to continue our series with Jake Bruce (DeepMind) who will be talking about their paper Genie: Generative Interactive Environments.

The talk will be hybrid and you will be able to listen and ask questions both in person or in remote. After the talk there will be a catered networking session. Please note that the link to the session will be available at the end of the confirmation email, where it says "Additional information".

Abstract
We introduce Genie, the first generative interactive environment trained in an unsupervised manner from unlabelled Internet videos. The model can be prompted to generate an endless variety of action-controllable virtual worlds described through text, synthetic images, photographs, and even sketches. At 11B parameters, Genie can be considered a foundation world model. It is comprised of a spatiotemporal video tokenizer, an autoregressive dynamics model, and a simple and scalable latent action model. Genie enables users to act in the generated environments on a frame-by-frame basis despite training without any ground-truth action labels or other domain-specific requirements typically found in the world model literature. Further the resulting learned latent action space facilitates training agents to imitate behaviors from unseen videos, opening the path for training generalist agents of the future.

About the speaker

Jake has a background in deep learning for robot navigation from Queensland University of Technology in Brisbane, Australia. He has been working on reinforcement learning and large-scale deep learning at Google DeepMind since 2018, and has been involved in projects on imitation learning in NetHack, generalist agents via Gato, and most recently generative world models in Genie.

Sponsors

This event is sponsored by InstaDeep and Google DeepMind