Member of Technical Staff Post Training

You will own multimodal post-training from data curation through deployment. You will advance fine-tuning, preference optimization, distillation, safety tuning, and evaluation; build personalization capabilities and high-throughput infrastructure; and identify and close quality and alignment gaps through research and engineering.

Responsibilities

• Own the post-training pipeline from data curation through deployment
• Advance fine-tuning, preference optimization, distillation, safety tuning, and evaluation
• Develop multimodal post-training capabilities
• Build personalization and customization capabilities
• Design and maintain high-throughput fine-tuning and evaluation infrastructure
• Identify and close model-quality and alignment gaps

Requirements

• Experience owning post-training for a frontier generative model through release
• Experience with reward modeling, preference learning, RLHF, RLAIF, and personalization
• Experience across text-to-image, image editing, multi-reference, or video models
• PyTorch proficiency
• Measurable model-quality improvements delivered to users

Benefits

• Equity
• Reasonable travel costs covered