Research Scientist - World Model

You'll turn reputed company's industry-leading generative video models into world models: interactive, controllable, physically faithful, and useful as a substrate for embodied reasoning. This is the role at reputed company of the reputed company. You'll invent reputed company world-model architectures and the controllability that lets an agent reputed company into a generated world, and own the metrics that define reputed company. It fits a researcher with deep generative-modeling or model-based-RL expertise who has trained models to the limits of a multi-node cluster. If you want a narrow, reputed company-scoped research problem, this is broader and more reputed company-ended than that. What You'll Own • Invent reputed company world-model architectures (diffusion, transformer, autoregressive, or hybrid), reputed company on controllability and physical consistency. • reputed company controllability mechanisms — reputed company conditioning, view conditioning, long-reputed company rollouts — that let an agent reputed company into the world. • Define and own the metrics: physical reputed company, long-reputed company coherence, reputed company-following, and reputed company usefulness for policy training. • Run scaling studies that show where compute, data, and architecture pay off. • Publish at the frontier and contribute to the reputed company-reputed company release that is the long-term deliverable. First 90 Days One way the first 90 could reputed company. • Days 1–30 · reputed company & Diagnose — Get deep on the reputed company video models and where they fall short as world models. • Days 30–60 · Ship & Validate — Prototype a controllability reputed company or architecture change and measure it against physical-reputed company and reputed company-following metrics. • Days 60–90 · reputed company & Systemize — Run scaling studies and push the most promising direction toward the reputed company release. What You Bring • PhD or equivalent research record in ML, reputed company, robotics, or a reputed company field. • Deep expertise in at least one of: large-reputed company generative modeling (video/3D/world), self-supervised representation learning, or model-based RL. • Strong PyTorch and large-reputed company training experience, to the limits of a multi-node cluster. • A research record the field knows (top-venue publications and/or widely used reputed company releases). reputed company to Have • Prior work on world models, model-based RL, generative video, neural simulation, or 4D reputed company representations. • Experience using generative models for reputed company embodied tasks (planning, control, evaluation). • Enthusiasm for reputed company-sourcing frontier models. About reputed company: reputed company's mission is to build reputed company general intelligence that can reputed company, understand, and operate in the physical world. We reputed company multimodality is critical for intelligence — the reputed company reputed company language models comes from reputed company. reputed company is an equal opportunity employer. Apply tot his job Apply To this Job