Researcher, Training - London

Details

OpenAI's Training team produces the large language models used across research and products. The team combines deep research on architectures and optimization with engineering work to improve efficiency and capability of future models.

Responsibilities

• Design, prototype, and scale up new architectures to improve model intelligence.
• Execute and analyze experiments autonomously and collaboratively.
• Study, debug, and optimize both model performance and computational performance.
• Contribute to training and inference infrastructure.

Requirements

• Deep understanding of large language model (LLM) architectures and transformer modifications for efficiency.
• Sophisticated understanding of model inference and a hands-on empirical approach to modelling and evaluation.
• Experience contributing to major LLM training runs and independently evaluating and improving deep learning architectures.
• Interest/experience in areas such as architecture design, long-context and efficient attention, optimization, and the science of scaling.
• Ability to design evaluations, debug regressions, and identify performance or computational bottlenecks.
• Motivation to safely deploy LLMs in real-world settings.

Workplace & Location

• This role is based in OpenAI's London office. The team follows a hybrid schedule (three days a week in the office; option to work from home on Thursdays and Fridays).
• The company is not considering remote-only applications for this role.
• The company offers relocation support for new employees joining in person.

Benefits & Compensation

• Salary range: £170,000 - £445,000 (total compensation includes competitive salary, equity, and benefits).
• Private medical insurance covering 100% of premiums for employees and their dependents.
• Pension plan with 4% employer contribution.
• 52 weeks maternity leave and 20 weeks parental leave.
• Unlimited time off.
• Annual learning & development stipend (£1,200 per year).