Summary

This senior research scientist role focuses on developing efficient model architectures and methods for attention, memory, reasoning, and adaptation. The stack includes Python, transformers, sparse and linear attention, model distillation, distributed training, and efficient inference; the working arrangement is not specified.

Responsibilities

• Formulate original research questions and experimental programs
• Design and evaluate architectural changes at scale
• Develop methods reducing training and inference costs
• Collaborate with engineers on efficient implementations
• Publish research and contribute to open-source models
• Mentor researchers and shape research direction