Cohere
Member of Technical Staff - Post-Training (Reinforcement Learning & Generalization)
Europe +2 · Remote · Staff · Full-time · Posted Oct 14, 2026
- Location
- Europe
New York
Toronto
About the job
Train Cohere’s large language models with reinforcement learning and develop, evaluate, and productionize methods that improve generalization beyond training data.
What you'll do
- Design, run, and analyze large-scale RL post-training experiments.
- Develop methods and held-out evaluations for out-of-distribution generalization, including detecting reward hacking and benchmark overfitting.
- Diagnose training failures and turn successful research into reproducible production training recipes.
What they ask for
- Strong research and software engineering skills, with hands-on experience training LLMs.
- Experience applying reinforcement learning to LLMs.
- Track record in generalization, robustness, or distribution shift through publications or shipped models.
- Rigorous experimental practice and strong understanding of data quality.
Nice to have
- Experience with distributed training, such as JAX, PyTorch, FSDP, or Megatron.
Pay
CA$295K – CA$535K CAD base salary
Tools
LLMs, PPO, GRPO, RLHF, RLVR, JAX, PyTorch, FSDP
AI focus
Reinforcement-learning post-training of LLMs, with research focused on out-of-distribution generalization, evaluation, and production training methods.
Model-written summary · the original posting is authoritative