Tech Stack
Tag name is followed by "@" symbol and proficiency level value.
About proficiency levels:
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
Debugging
Deep Learning @ 3
LLM @ 6
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
Details
OpenAI's Training team produces the large language models that power its research and products. The team combines research into architecture and optimization techniques with long-term efforts to improve the efficiency and capabilities of future model generations. It integrates these techniques into model artifacts used across the company and ensures that the models are world-class.
As a member of the Training team, you will advance large language model development for OpenAI's flagship models by improving intelligence, efficiency, and capabilities. Relevant areas include architecture design, long-context and efficient attention, optimization, and the science of scaling.
Ideal candidates have a deep understanding of LLM architectures, a sophisticated understanding of model inference, and a hands-on, empirical approach. The role involves developing creative breakthroughs, strengthening baselines, designing evaluations, debugging regressions, and identifying performance bottlenecks.
Responsibilities
- Design, prototype, and scale new architectures to improve model intelligence.
- Execute and analyze experiments autonomously and collaboratively.
- Study, debug, and optimize model and computational performance.
- Contribute to training and inference infrastructure.
Requirements
- Experience contributing to major LLM training runs.
- Ability to thoroughly evaluate and improve deep learning architectures in a self-directed manner.
- Motivation to safely deploy LLMs in the real world.
- Strong knowledge of state-of-the-art transformer modifications for efficiency.
Workplace & Location
This role is based in OpenAI's London office and is not available remotely. The hybrid schedule requires three days per week in the office, with the option to work from home on Thursdays and Fridays.
Benefits
- Relocation support for employees joining in person.
- Private medical insurance covering 100% of premiums for employees and dependents.
- Pension plan with a 4% employer contribution.
- 52 weeks of maternity leave and 20 weeks of parental leave.
- Unlimited time off.
- Annual learning and development stipend of £1,200.
- Competitive salary, equity, and benefits.
OpenAI is an equal opportunity employer committed to providing reasonable accommodations to applicants with disabilities.