ConceptAI & ML
World model
A component in an AI architecture that predicts future world states resulting from an imagined action sequence.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Opens up
Concepts that name this one as a prerequisite.
Lecturers
7Each in their own terms, from the easiest level.
Demis Hassabis
A system that has a real understanding of the real world and how interactions and physics work within it.Steven Feng
Structured representations of environments used to predict future states rather than just text tokens.MIT Video Productions
A learned system that internally represents scene geometry, materials, lighting, and dynamics to directly predict images from state, viewpoint, and time.Chelsea Finn
Models that capture rich knowledge about the world and can be used for video generation and planning.Jon Kleinberg
An internal representation within an agent or language model that captures locations, routes, transportation options, and the state of the world to solve tasks like navigation.John Langford
A model composed of a belief state and a transition model over those belief states.
Lectures that assume it
2Connections
18The lectures that define it or assume it, and the concepts next to it.
Concepts9
Show 3 moreShow fewer
Lectures9
- Frontiers of Deep Reinforcement Learning
- Venturing out into the world with AI
- Objective-Driven AI: World Models that Learn, Remember, Reason, and Plan
- Accelerating Scientific Discovery with Artificial Intelligence
- Overview of Transformers (Stanford CS25: Transformers United V6)
- An Occam's Razor Principle for Transformers?