Jobiglo

No results.

This job is no longer available

This job expired on 24/10/2026. It no longer accepts applications.

Midtraining Research Engineer

periodic-labs · Menlo Park

Mid 🇬🇧 English
self-distillation on-policy distillation scaling laws distributed training

Job description

About the role

We are an AI and physical sciences company building state‑of‑the‑art models to accelerate breakthroughs in materials, energy, and beyond. As a Midtraining Research Engineer you will improve the scientific reasoning of frontier models by curating data, generating synthetic examples, building evaluations, and running large‑scale training experiments.

Key responsibilities

  • Identify, process, and curate novel scientific data sources for large‑scale model training.
  • Generate high‑quality synthetic data to fill gaps in scientific knowledge.
  • Design and implement evaluations that correlate with downstream scientific task performance.
  • Develop and apply self‑distillation and on‑policy distillation techniques to boost model capability.
  • Design and execute large‑scale training experiments, collaborating with super‑compute engineers to scale across thousands of GPUs.
  • Build internal tools to investigate how data choices affect model intelligence.

Required profile

  • Hands‑on experience training large language models on curated mixes of trillions of tokens.
  • Background on an evals team supporting large production training runs.
  • Practical use of self‑distillation, on‑policy distillation, or similar methods in real pipelines.
  • Understanding of scaling laws and compute‑optimal hyperparameters.
  • Comfort working across data, evaluation, and training infrastructure.
  • Bonus: experience optimizing throughput and reliability for distributed training, or AI for science domains such as protein or materials.

Required skills

  • LLM training at trillion‑token scale
  • Self‑distillation and on‑policy distillation techniques
  • Scaling laws and compute‑optimal hyperparameter tuning
  • Large‑scale distributed training on GPU clusters
  • Synthetic data generation for scientific tasks
  • Evaluation framework development for scientific reasoning

What we offer

  • Base salary $250,000–$350,000 per year plus equity
  • Visa sponsorship and full assistance with immigration
  • Opportunity to work at the frontier of AI‑driven scientific discovery
  • Collaborative environment with physicists, chemists, and RL researchers

Questions fréquentes

Le salaire n'est pas communiqué publiquement par le recruteur. Vous pouvez postuler et négocier directement avec periodic-labs.
Cliquez sur "Postuler maintenant" en haut de la page. Vous pouvez importer votre CV en 1 clic — Jobiglo extrait automatiquement vos informations et postule pour vous.
Source : ats:ashby

Why are you reporting this job?

Thank you for your report. We will review this job.

Explore further

Salaries, guides and searches in the United States.

A question about this job?

Ask it here: you will get the full job summary by e-mail, right away.

Published 2 months ago

17 views · 0 interested

Boost your chances

Upload your CV — we will match you with relevant openings.

Analyzing your CV...

periodic-labs

Menlo Park