Genie 2
Google DeepMind
Large-scale foundation world model for generating action-controllable interactive environments from visual prompts.
Controllable world generation
Not a robot policy, but a useful trend marker for the world-model side of embodied AI.
A model hub link is a declaration. Files, loadability, evaluation, and deployment are scored separately.
Model decision scorecard
Use-case scores and evidence confidence are separate; Unknown is not treated as failure.
Code, weights, and checkpoints
UnknownArtifact availability is unknown.
Loading and training reproducibility
UnknownNo verified loading configuration is available.
Training data requirements
UsefulRequired signal categories are declared; exact tensor and action interfaces still need verification.
Evaluation evidence
UsefulEvaluation focus is declared, but metrics are not independently verified.
Deployment readiness
UnknownHardware, latency, dependencies, and runtime loading are not yet verified.
Artifact facts and provenance
No metadata-verified artifact facts yet. Source links remain declarations only.
Loop signal demand
Signals this model family needs for training, evaluation, or failure mining.
Observation / ego video
observation · Action-controllable video or environment interaction traces · Consistent observations over time with enough geometry and dynamics
Language intent / task phase
Needs task intent, instructions, keysteps, or phase labels to connect behavior to goals.
Action / robot state
actions · Action-controllable video or environment interaction traces · Physical plausibility under action changes
Future state / dynamics
future state · Consistent observations over time with enough geometry and dynamics · Controllable world generation
Feedback / correction / failure
feedback/failure
Sim-real / embodiment metadata
sim-real
Evaluation focus
- Controllable world generation
- Temporal persistence
- Physical plausibility under action changes
Missing critical loop signals
Core signal demands are represented. Check quality, alignment, and access constraints.
Related catalog datasets
Ego-Realm
Small egocentric sample set for manipulation domains
Exact action dimensions, control frequency, normalization, and camera mapping require interface verification.
Xperience-10M
Multimodal human experience for embodied AI
Exact action dimensions, control frequency, normalization, and camera mapping require interface verification.
Ego-1K
Multiview egocentric scene reconstruction
Exact action dimensions, control frequency, normalization, and camera mapping require interface verification.
OpenBot notes
- Not a robot policy, but a useful trend marker for the world-model side of embodied AI.
- Supports OpenBot's argument that data should preserve actions and feedback, not only frames.