🛠 Tech Stack

💼 About This Role

You'll train frontier models to become highly knowledgeable scientific experts, developing methods for synthetic data generation and continual learning at scale. You'll collaborate closely with RL researchers, physicists, and chemists to create evals that guide scientific data curation.

🎯 What You'll Do

• Train frontier models to be scientific experts for RL
• Develop synthetic data generation and distillation methods
• Create evals to guide scientific data curation
• Scale LLM training to thousands of GPUs

📋 Requirements

• Experience training LLMs on curated trillion-token mixes
• Understanding of scaling laws and compute-optimal hyperparameters
• Experience generating billions of tokens of synthetic data
• Ability to build evals correlating with downstream tasks

✨ Nice to Have

• Familiarity with RL for LLMs
• Background in physics or chemistry
• Experience with high-performance computing

🎁 Benefits & Perks

• 🚀 Well-funded, fast-growing startup
• 🧑‍💻 Remote-friendly culture
• 📈 Ownership culture with no bureaucracy
• 🛠️ Access to supercomputing resources