Isomorphic Labs · 201–500 people · Series A
Research Engineer (LLM Performance), London
- London
- Hybrid · 3 days in office
- Mid-level
Join the model performance and scaling team to implement and optimize LLM post-training methods on frontier models, working with JAX or PyTorch, parallelism strategies and GPU architectures. You will diagnose bottlenecks in distributed training and inference, deploy low-precision methods, and collaborate with research teams to turn new methods into production systems for drug design.
JAX · PyTorch · NCCL +5