Post-Training Engineer for Production Models

Anthropic Limited; New York (NY); On-site; USD 350,000 - 500,000; Full time

Benefits offered by this job

Equity donation matching

Generous vacation

Parental leave

Flexible working hours

Job summary

Anthropic is seeking a Research Engineer on the Post-Training team to train base models through the full post-training stack, delivering production Claude models used by users. You will implement post-training techniques such as Constitutional AI and RLHF, scale pipelines, and build tools to measure model performance, collaborating with research and engineering teams.

Location-based hybrid policy may apply; accessibility to offices and visa sponsorship are considered as part of the role, with a

Qualifications

• Bachelor’s degree or equivalent in a relevant field.
• Strong Python and deep learning framework proficiency.
• Experience with training, fine-tuning, or evaluating large language models.

Responsibilities

• Implement and optimize post-training techniques at scale on frontier models.
• Conduct research to develop and optimize post-training recipes that directly improve production model quality.
• Design, build, and run robust pipelines for model fine-tuning and evaluation.
• Develop tools to measure and improve model performance across various dimensions.
• Collaborate with research teams to translate emerging techniques into production-ready implementations.
• Debug complex issues in training pipelines and model behavior.
• Help establish best practices for reliable, reproducible model post-training.

Skills

Python

Deep learning

Distributed computing

LLMs

Research

Education

Bachelor's degree

Tools

PyTorch

Job description

Anthropic is seeking a Research Engineer on the Post-Training team to train base models through the full post-training stack, delivering production Claude models used by users. You will implement post-training techniques such as Constitutional AI and RLHF, scale pipelines, and build tools to measure model performance, collaborating with research and engineering teams.

Location-based hybrid policy may apply; accessibility to offices and visa sponsorship are considered as part of the role, with a