Jack Parker-Holder, DeepMind: Open-endedness, evolving agents and environments, online adaptation, and offline learning
Generally Intelligent
00:00
QDax, Is It a Lab?
This was actually a spotlight, our workshop, the Allea workshop. Now you can run these party diversity algorithms just on a single GPU. And it's super fast. I think that people in RL might start realizing this and realize that there's a lot of overlap. Ultimately alpha style, for example, is still one of the most exciting large scale RL projects using many ideas from this community.
Play episode from 01:33:46
Transcript
