Research Engineer/Scientist (all levels), Efficient Models

Responsibilities

• Develop efficient algorithms and architectures for large-scale generative and multimodal models.
• Utilize techniques such as step distillation, CFG distillation, and quantization to improve model efficiency for image and video generation.
• Advance scalable generative modeling approaches, including diffusion and autoregressive models, with a focus on acceleration.
• Design and implement methods for transferring capabilities from foundation models into smaller, more efficient models.
• Work on distillation frameworks, model acceleration, and hardware-efficient inference.

Requirements

• B.S. in Computer Science or a related field, or equivalent experience.
• Expertise in efficient models with a deep understanding of computational bottlenecks and acceleration methods.
• Proficiency in training generative AI or LLM models using frameworks such as PyTorch or JAX.
• Strong communication and collaboration skills in fast-paced environments.

Preferred Qualifications

• Ph.D. in GenAI, MLSys, or equivalent experience.
• Extensive research experience in broad GenAI, MLSys, or LLM areas.
• Proven experience in image/video generation, model compression (quantization, distillation), or efficient architectures (MoE, window attention).
• Experience with reinforcement learning training methods such as RLHF, DPO, or GRPO.

Benefits

• Medical, dental, and vision insurance.
• 401(k) savings plan with company match.
• Paid parental leave and wellbeing benefits.
• Paid time off including 10 holidays, 10 sick days, and 17 days of personal time.

About the Company

ByteDance's mission is to inspire creativity and enrich life. With a suite of more than a dozen products, including TikTok, Lemon8, CapCut, and Douyin, ByteDance makes it easier and more fun for people to connect, consume, and create content globally.