Multimodal Games Research Scientist

Tencent•Palo Alto, CA

1d•Onsite

About The Position

At Tencent, we believe the next generation of AI will fundamentally reshape how people create, interact, and experience digital worlds. As part of our ongoing investment in frontier AI technologies, we are building cutting-edge generative AI systems for interactive game worlds and immersive content creation. We are looking for exceptional researchers and engineers who are passionate about advancing multimodal intelligence and world models at scale.

Requirements

MS or PhD in Computer Science, Artificial Intelligence, Mathematics, or related fields; PhD preferred.
Strong programming and engineering skills, with hands-on experience in PyTorch or other modern deep learning frameworks.
Experience reproducing and extending state-of-the-art research and translating research ideas into scalable systems.
Expertise in one or more of the following areas: Generative models such as Diffusion Models, GANs, VAEs, or Flow-based Models; Image or video generation; Large-scale model pretraining, fine-tuning, or post-training on GPU clusters; Multimodal foundation models such as CLIP, LLaVA, or Qwen.

Nice To Haves

Publications at top-tier conferences such as SIGGRAPH, CVPR, ICCV, or NeurIPS are a strong plus.

Responsibilities

Research and develop core algorithms for Generative Game Engine technologies, including but not limited to: Video generation, Video prediction, Video editing, Related multimodal generative modeling tasks.
Explore next-generation video generation technologies for interactive game-world scenarios. Design and build world models with capabilities such as: Interactivity, Editability, Physical consistency, Spatiotemporal consistency.