Researcher, Training - London

United States Digital Space LLC; Greater London; Hybrid; GBP 70,000 - 90,000; Full time

Benefits offered by this job

Relocation support

Hybrid work schedule

Job summary

United States Digital Space LLC is seeking a LLM Training Engineer to enhance the intelligence and efficiency of flagship models. This role involves designing new architectures and optimizing model performance. The ideal candidate should possess substantial experience in LLM advancement and a proactive approach to problem-solving.

This position requires on-site presence in London with a hybrid work setup—three days a week in-office while allowing remote work on Thursdays and Fridays. Relocation support is also provided.

Qualifications

• Experience landing contributions to major LLM training runs.
• Ability to thoroughly evaluate and improve deep learning architectures.
• Motivation for safely deploying LLMs in the real world.

Responsibilities

• Design, prototype, and scale up new architectures to improve model intelligence.
• Execute and analyze experiments autonomously and collaboratively.
• Study, debug, and optimize model performance.

Skills

Experience in LLM training

Deep learning architecture evaluation

Transformer modifications

Job description

About the Team

the company's Training team is responsible for producing the large language models that power our research, our products, and ultimately bring us closer to AGI. Achieving this goal requires combining deep research into improving our current architecture and optimization techniques, alongside long‑term bets aimed at improving the efficiency and capability of future generations of models. We are responsible for integrating these techniques and producing model artifacts used by the rest of the company, and ensuring that these models are world‑class in every respect.

About the Role

As a member of the training team, you will push the frontier of LLM development for the company's flagship models, enhancing intelligence, efficiency, and adding new capabilities. Relevant interests may include areas such as 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. A good fit for this role will be equally happy coming up with a creative breakthrough, investing in strengthening a baseline, designing an eval, debugging a thorny regression, or tracking down a bottleneck.

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

Qualifications

• Experience landing contributions to major LLM training runs
• Ability to thoroughly evaluate and improve deep learning architectures in a self‑directed fashion
• Motivation for safely deploying LLMs in the real world
• Well‑versed in state‑of‑the‑art transformer modifications for efficiency

Workplace & Location

This role is based in London, and we require on‑site presence; remote work is not an option. We offer relocation support and a hybrid schedule of three days a week in the office, with option to work from home on Thursdays and Fridays.

Equal Opportunity

We are an equal opportunity employer, and we do not discriminate on the basis of race, religion, color, national origin, sex, sexual orientation, age, veteran status, disability, genetic information, or other legally protected characteristic.