Research Engineer (LLM Performance), London
Isomorphic Labs · London
ResearchCompany career page
The apply link worked recently (not employer-confirmed).
Role summary
Isomorphic Labs seeks a Research Engineer to optimize large language models for drug discovery applications. The role involves implementing post-training methods, translating research into production systems, and solving performance challenges in distributed training and inference.
What they're looking for
- Significant experience scaling distributed LLM training systems
- Proficiency with JAX or PyTorch deep learning frameworks
- Knowledge of parallelism strategies and collective communication libraries
- Understanding of GPU architectures and performance optimization concepts
- Experience with supervised fine-tuning and reinforcement learning methods
What you'll be doing
- Implement and optimize LLM post-training techniques at scale on frontier models
- Collaborate with research teams to translate new methods into production systems
- Identify and resolve performance bottlenecks in distributed training and inference
- Evaluate and deploy frameworks for fine-tuning, reinforcement learning, and model evaluation
- Deploy low-precision methods to balance model performance with computational efficiency
Privacy & handoff
This role links to Isomorphic Labs's own application system. RoleLantern records your interest and checks the link, but the application happens on the employer's site and status updates may not come back to us. Clicking apply does not mark you as applied automatically.