AI Safety Post-Training Research Scientist

JRNLClub; New York (NY); On-site; USD 216,000 - 270,000; Full time

Benefits offered by this job

Equity

Learning stipend

Generous PTO

Commuter stipend

Health benefits

Job summary

Scale Labs, Scale AI's research arm, is seeking a Research Scientist focused on Safety Post-Training to develop and apply post-training methods and interpretability techniques that improve frontier AI systems.

You will design post-training pipelines, build interpretability-informed evaluations, and collaborate with policymakers, engineers, and researchers to translate findings into safety standards and benchmarks.

Qualifications

• Experience developing post-training pipelines to study safety and alignment in frontier models.
• Experience with interpretability evaluations to reveal unsafe or undesirable model behavior.
• 3+ years of ML research experience addressing sophisticated ML problems.
• Strong written and verbal communication for cross-functional collaboration.

Responsibilities

• Design and run post-training pipelines to study how training choices affect model safety, robustness, and alignment properties.
• Develop interpretability-informed evaluations that reveal unsafe or undesirable model behaviors, guiding mitigations.
• Collaborate with policymakers, engineers, and researchers to translate findings into safety standards and benchmarks.

Skills

Post-training methods

Interpretability

RLHF

Policy collaboration

Research publishing

Job description

Scale Labs, Scale AI's research arm, is seeking a Research Scientist focused on Safety Post-Training to develop and apply post-training methods and interpretability techniques that improve frontier AI systems.

You will design post-training pipelines, build interpretability-informed evaluations, and collaborate with policymakers, engineers, and researchers to translate findings into safety standards and benchmarks.