Research Engineer, Post-Training LLMs & RL Environments

Preference Model, Inc.; San Francisco (CA); On-site; USD 190,000 - 270,000; Full time

Benefits offered by this job

Equity compensation

Health benefits

401K match

Lunch onsite

Weekly snacks

Visa sponsorship

Relocation support

Job summary

Preference Model, Inc. is seeking machine learning Research Engineers or Research Scientists to push the frontier of post-training on large language models, blending research and engineering to implement novel approaches and shape research directions.

You will train and evaluate models on our proprietary RL environments, architect scalable training infrastructure, design training environments, and optimize end-to-end pipelines to accelerate discovery.

Qualifications

• Experience in ML research and development with focus on language models.
• Strong programming skills in Python and ML frameworks.
• Familiarity with RL training frameworks and distributed training.
• Experience building and operating ML infrastructure at scale.
• Ability to translate research ideas into running code.
• Comfort with evaluating model outputs and reward signals.

Responsibilities

• Train and evaluate models on proprietary RL environments to validate data quality and task coverage.
• Architect and optimize RL training infrastructure and experiment management to scale workflows.
• Design, implement, and test training environments, evaluations, and methodologies for RL agents.
• Profile and optimize end-to-end training runs to maximize throughput and shorten iteration cycles.

Skills

Machine learning

Python

PyTorch

JAX

RL frameworks

Distributed systems

Tools

Verl

OpenRLHF

ML infrastructure

Job description

Preference Model, Inc. is seeking machine learning Research Engineers or Research Scientists to push the frontier of post-training on large language models, blending research and engineering to implement novel approaches and shape research directions.

You will train and evaluate models on our proprietary RL environments, architect scalable training infrastructure, design training environments, and optimize end-to-end pipelines to accelerate discovery.