NVIDIA
Senior AI Researcher - World Foundation Models
$184K–$356K
US, CA, Remote +2 · Senior · Full-time · Posted Jun 23, 2026
- Location
- US, CA, Remote
US, OR, Remote
US, WA, Remote - Experience
- 8+ years of applied research and/or industry experience in vision, graphics, or adjacent ML domains or similar area; 3+ years of direct generative-model experience.
- Education
- PhD in Computer Science, Graphics, Computer Engineering, or a closely related field (or equivalent experience).
About the job
Develop and validate world foundation models for high-fidelity, temporally stable video and world generation, focusing on human appearance, motion, and action understanding, and deliver improvements as production-ready models and tools.
What you'll do
- Research, implement, and evaluate model architecture and algorithm improvements for video generation and human-centric quality.
- Improve model training and inference efficiency, and develop training objectives for sim-to-real and real-to-sim generalization.
- Create domain-specific benchmarks and translate results into training code, production-grade checkpoints, integrations, and demos.
What they ask for
- PhD in Computer Science, Graphics, Computer Engineering, or a closely related field (or equivalent experience).
- 3+ years of direct experience designing, training, and evaluating generative models for image/video/audio.
- Hands-on experience improving generative models for perceptual quality and temporal stability, especially human generation.
- Advanced proficiency in Python, PyTorch, C++, and CUDA.
Nice to have
- Publications in top conferences such as NeurIPS, CVPR, or ICLR, with evidence of impact on model quality or robustness.
- Experience using agentic workflows and AI coding companions to accelerate research and production development.
Pay
184,000 USD - 287,500 USD (base salary); 224,000 USD - 356,500 USD (base salary)
Tools
Python, PyTorch, C++, CUDA
AI focus
Research and improve world foundation models for video and world generation, including spatial multimodal modeling, modality alignment, flow-based or diffusion-based generation, neural-rendering-inspired representations, and model training and evaluation. AI coding companions and agentic workflows are listed as preferred experience.
Model-written summary · the original posting is authoritative