Senior ML Researcher, Post-Training & RL for LLMs

Wwshemi; United States; Remote; USD 140,000 - 200,000; Full time

Benefits offered by this job

Weekly lunch stipend

Health and dental benefits

RRSP matching / 401K / Pension

Parental leave top-up

Enrichment benefits

Education stipend

Vacation 6 weeks

Travel budget between offices

Home office stipend

Job summary

Cohere is the leading security-first enterprise AI company building foundation models and end-to-end products to solve real-world business problems. This role is for a Member of Technical Staff in Post-Training to advance RL-based training of large language models.

You will design training recipes, environments, and evaluations, run large-scale RL experiments (RLHF, RL from verifiable rewards), and turn insights into production-ready training pipelines in collaboration with post-training data,

Qualifications

• Strong research and software engineering skills, with hands-on experience training LLMs.
• Experience applying reinforcement learning to LLMs (e.g., PPO, GRPO, RLHF, RLVR), ideally at scale.
• Track record of work on generalization, robustness, or distribution shift, through publications or shipped models.
• Rigorous experimental practice: clean ablations, careful controls, and sound conclusions from noisy results.
• Experience with distributed training (e.g., JAX, PyTorch, FSDP, Megatron) is a plus.
• Excellent communication skills.
• Deep understanding of data quality principles and best practices: demonstrates willingness to dive deep into data while possessing strong intuition for assessing the quality of datasets and anticipating how different data characteristics will impact model training outcomes.]

Responsibilities

• Design, run, and analyse large-scale RL post-training experiments (e.g., RLHF, RL from verifiable rewards).
• Research and develop methods that improve out-of-distribution generalization, such as reward design, environment and task diversity, curricula, and regularization.
• Build held-out evaluations that measure generalization to unseen tasks, domains, and formats, and that catch reward hacking and benchmark overfitting.
• Diagnose training dynamics and failure modes, including instability, reward over-optimization, and capability regressions.
• Turn successful research into robust, reproducible training recipes for production models.

Skills

LLM training

Reinforcement learning

Distributed training

Research background

Communication skills

Software engineering

Tools

JAX

PyTorch

FSDP

Megatron

Job description

Cohere is the leading security-first enterprise AI company building foundation models and end-to-end products to solve real-world business problems. This role is for a Member of Technical Staff in Post-Training to advance RL-based training of large language models.

You will design training recipes, environments, and evaluations, run large-scale RL experiments (RLHF, RL from verifiable rewards), and turn insights into production-ready training pipelines in collaboration with post-training data,