Job Overview
Location
Los Angeles, CA, New York, NY, SF Bay Area, CA, Remote, US, International, US, Remote, International, Singapore, SGP
Job Type
Full-time
Category
Data Science
Date Posted
June 3, 2026
Full Job Description
đź“‹ Description
- • Invent next-generation world model architectures, including diffusion, transformer, autoregressive, or hybrid models, with a focus on controllability and physical consistency.
- • Develop controllability mechanisms such as action conditioning, view conditioning, and long-horizon rollouts to enable agents to interact with generated worlds.
- • Define and own key metrics for evaluating world models, including physical fidelity, long-horizon coherence, action-following accuracy, and downstream usefulness for policy training.
- • Conduct scaling studies to determine optimal allocations of compute, data, and architectural resources for model performance and efficiency.
- • Publish research at top-tier venues and contribute to open-source releases that serve as the long-term deliverable for the World Models team.
- • Work at the intersection of generative modeling, computer vision, and embodied reasoning to build systems that can generate, understand, and operate in the physical world.
- • Collaborate with cross-functional teams to integrate world models into broader AI systems capable of vision, interaction, and change.
- • Train and optimize large-scale generative models that push the limits of multi-node cluster infrastructure using PyTorch.
- • Advance the state of the art in generative video, neural simulation, and 4D scene representations to enable physically faithful and interactive environments.
- • Translate research into practical applications by using generative models for downstream embodied tasks such as planning, control, and evaluation.
- • Maintain a strong research presence through top-venue publications and widely-used open-source contributions that shape the field.
- • Embrace open-sourcing frontier models as a core part of the team’s mission to advance collective understanding of world modeling.
🎯 Requirements
- • 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).
🏖️ Benefits
- • Base pay range of $250,000 – $450,000 per year.
- • Opportunity to publish at the frontier and contribute to open-source releases of cutting-edge models.
- • Work on mission-critical research to build unified general intelligence that can generate, understand, and operate in the physical world.
- • Flexible location: eligible for roles in Los Angeles, CA; New York, NY; SF Bay Area, CA; or remote in the US or internationally.
Skills & Technologies
About Luma Labs, Inc.
Luma Labs is a technology company focused on developing advanced AI-powered tools for 3D content creation. Their flagship product enables users to capture real-world objects and environments and transform them into high-fidelity 3D models using just a smartphone. This technology serves various industries, including gaming, augmented reality (AR), virtual reality (VR), and e-commerce, by democratizing the creation of immersive digital assets. Luma Labs aims to make 3D scanning and modeling accessible to a broader audience, accelerating the development of the metaverse and other spatial computing applications.
Subscribe to the weekly newsletter for similar remote roles and curated hiring updates.
Newsletter
Weekly remote jobs and featured talent.
No spam. Only curated remote roles and product updates. You can unsubscribe anytime.
