What You'll Do
Work on the forefront of 3D generative AI by designing and refining deep learning models that generate low-poly meshes or high-fidelity textures. Explore autoregressive and discrete diffusion frameworks to push the boundaries of what generative systems can create.
Process and structure 3D mesh data for model input, including developing tokenization strategies that allow neural networks to effectively interpret geometric information. Focus on architectural improvements that enhance training efficiency, optimizing both CPU and GPU utilization to reduce bottlenecks.
Advance texture generation using diffusion and autoregressive methods, aiming for outputs with sharp detail and strong alignment to underlying geometry. Deliver production-ready 3D assets with full PBR material stacks that respond accurately to lighting conditions in real-time environments.
Requirements
- Available for a full-time, 12-week (or longer) internship
- Planning to pursue a full-time role at the company post-graduation, ideally between September 2026 and September 2027
- Currently enrolled in an undergraduate, master’s, or PhD program
- Solid grasp of transformer-based architectures and their application beyond sequential data
- Proficient in Python and experienced with tensor-level programming (e.g., PyTorch)
- Proven ability to turn abstract research goals into actionable technical plans
- Strong communication skills for articulating complex ideas clearly
- Self-motivated, collaborative, and comfortable working in a fast-moving research environment
- Belief in the power of scaling laws to drive model performance
- Curiosity about 3D geometry, rendering, and generative systems
Benefits
Contribute to cutting-edge AI research that shapes how machines understand and generate 3D environments. Your work will directly influence a leading generative platform used by millions, from indie creators to enterprise teams.
Engage with a community of builders and artists who share and refine digital assets. See your innovations applied across diverse fields—game development, film, industrial design, robotics, and physical simulation.
Grow within a culture that values deep technical rigor, creative problem-solving, and shipping impactful research.