Hunyuan Multimodal Algorithm Researcher (Omni-Modal)

About the role

Tencent's Technology Engineering Group (TEG) is hiring an Omni-Modal Multimodal Algorithm Researcher in Palo Alto, California. The role covers research and development of omnimodal large models, from data design and foundational algorithms to model evaluation and downstream applications.

Key responsibilities

• Research and develop omnimodal large models, including designing and building training data, foundational model algorithms, pre-training/SFT/RL optimization, model evaluation and downstream scenario exploration.
• Analyze R&D challenges and model performance bottlenecks, devise first-principles solutions and speed up model iteration.
• Explore paradigms for omnimodal understanding and generation, and research next-generation model architectures.

Required profile

• Bachelor's degree or higher in Computer Science, Artificial Intelligence, Mathematics or a related field; graduate degrees are prioritized.
• Hands-on experience with large-scale multimodal data processing and high-quality data generation is highly preferred.
• Experience in cross-modal (e.g., audio-visual) research or publications in top-tier conferences is preferred.
• Strong learning agility, communication skills, teamwork and curiosity.

Required skills

• Solid foundation in deep learning algorithms and practical experience in large model development.
• Familiarity with Diffusion Models and Autoregressive Models is advantageous.
• Proficiency in deep learning network implementation, model tuning for training and inference, CPU/GPU acceleration and distributed training/inference optimization is a plus.
• Participation in ACM or NOI competitions is highly valued.

What we offer

• Expected base pay of $140,400 to $264,000 per year, depending on knowledge, skills and experience.
• Possible sign-on payment, relocation package and restricted stock units, evaluated case by case.
• Medical, dental, vision, life and disability benefits, and a 401(k) plan.
• Up to 15 to 25 vacation days, up to 13 holiday days and up to 10 paid sick days per year, depending on tenure.