Research Scientist - World Model
As a Research Scientist on the World Models team, you will invent next-generation world model architectures with a focus on controllability and physical consistency, develop controllability mechanisms, and define and own metrics for physical fidelity and action-following.
About the job
THE ROLE
This is the role at the center of the thesis. Luma already trains the strongest generative video models in the industry; the next step is turning those models into world models — interactive, controllable, physically faithful, and useful as a substrate for embodied reasoning. As a Research Scientist on the World Models team, you'll work on the next generation of generative models that can be rolled out as worlds.
WHAT YOU'LL DO
- Invent next-generation world model architectures — diffusion, transformer, autoregressive, or hybrid — with a particular focus on controllability and physical consistency.
- Develop controllability mechanisms that let an agent step into the world: action conditioning, view conditioning, long-horizon rollouts.
- Define and own the metrics: physical fidelity, long-horizon coherence, action-following, and downstream usefulness for policy training.
- Run scaling studies that tell us where compute, data, and architecture pay off.
- Publish at the frontier; contribute to the open-source release that is the long-term deliverable.
MINIMUM QUALIFICATIONS
- PhD or equivalent research record in ML, computer vision, robotics, or related fields.
- Deep expertise in at least one of: large-scale generative modeling (video/3D/world), self-supervised representation learning, model-based RL.
- Strong PyTorch and large-scale training experience — you've trained models that hit the limits of a multi-node cluster.
- A research record the field knows (top-venue publications and/or widely-used open releases).
PREFERRED
- Prior work on world models, model-based RL, generative video, neural simulation, or 4D scene representations.
- Experience using generative models for downstream embodied tasks (planning, control, evaluation).
- Excitement about open-sourcing frontier models.
Compensation
The base pay range for this role is $250,000 – $450,000 per year.
Skills
PyTorch, Generative Modeling, Machine Learning, Computer Vision, Robotics, Self-Supervised Learning, Model-Based Reinforcement Learning, Neural Simulation, 4D Scene Representations
Similar jobs
AI Research jobs
The Voleon Group
New York, NYConducts causal inference research for financial market prediction and portfolio optimization, developing and validating models from research through live trading. Requires Ph.D.-level coursework, strong causal inference and statistics expertise, mathematical ability, and production Python skills.
Descript
San Francisco, CA
Applied research scientists develop deep-learning and generative media systems for video, audio, and multimodal editing features that ship to millions of users. The role requires strong PyTorch or TensorFlow skills, rapid experimentation, and evidence of impactful research or production machine-learning work.
Baseten
San Francisco, CA
Build and ship agentic AI product experiences, internal automation, and customer-facing features across the stack. The role requires 5+ years of software engineering experience, hands-on experience with AI or LLM-powered products, Python proficiency, and strong autonomy.
OpenAI
San Francisco, CA
Conducts frontier AI research for health, developing and evaluating scalable training methods, models, and agents that improve medical reasoning, reliability, and real-world outcomes. Requires exceptional machine learning or biomedical AI research depth, hands-on coding and experimentation, and end-to-end ownership of ambiguous problems.
Mercor
San Francisco, CA
Leads the design, measurement, publication, and adoption of APEX benchmarks evaluating frontier models on economically valuable professional work. The role requires rigorous research judgment, strong coding and statistical skills, and excellent communication across technical, commercial, and research audiences.