Applied Science Intern - World Model

About the role#

As an Applied Science Intern, you will join the Valka.ai team to research and develop World Models. You will focus on the simulation and generation of sport and eSport matches, helping to bridge the gap between passive content and interactive digital experiences.

What you'll do#

• Explore the application of World Models for understanding, simulating, and generating sport or eSport matches, such as soccer or DOTA.
• Design, develop, and optimize AI video generation models, with a focus on experimenting with autoregressive architectures.
• Develop and implement algorithms for synthesizing sport matches.
• Collaborate with teams on large-scale video-action datasets, including the design and implementation of data-cleaning and pre-processing pipelines.
• Define validation strategies and implement custom evaluation metrics to compare synthetic gameplay against real-world data.

What you'll need#

• Currently pursuing a PhD, preferably based in the San Francisco area.
• A record of publication at top Computer Vision, AI, or Graphics venues such as CVPR, ICML, ICCV, Siggraph, or NeurIPS.
• Demonstrated hands-on experience building and running generative computer vision models, including GANs, DiT, or VAEs.
• A solid understanding of neural architectures and paradigms, including Transformers, Denoising Diffusion Models, RNNs, Sequence Models, and CNNs.
• Proficiency in Python and PyTorch.

Location & details#

• This is a remote, full-time internship position.
• The role is paid and operates on a rolling basis.