Foundation Model Post-Training Researcher & Engineer

Apple Inc.; Seattle (WA); Hybrid; USD 175,000 - 309,000; Full time

Benefits offered by this job

Medical and dental coverage

Retirement benefits

Employee stock plan

Job summary

Apple Inc. is seeking an ambitious ML Researcher to shape the future of large-scale foundation models. You will transform models into capable, privacy-forward assistants powering Apple products, tackling instruction following, tool use, and deep reasoning in a fast-growing team.

You will partner with pre-training and product teams, shaping architecture choices and evaluating model performance across practical tasks, while aligning with Apple's privacy and user-centric goals.

Qualifications

• Demonstrated expertise in deep learning with a focus on LLMs, post-training, or reinforcement learning.
• Proficient programming in Python and a major deep learning framework (JAX or PyTorch).
• Masters/PhD, or equivalent practical experience in Computer Science, ML, or related field.

Responsibilities

• Design and iterate on post-training strategies (including Reinforcement Learning) for model behaviors.
• Pioneer algorithms for preference optimization, model steering, and safety.
• Drive data strategy with high-quality human and synthetic data generation and automated filtering.
• Design evaluation methodologies for model helpfulness, factuality, and utility.
• Collaborate with pre-training and product teams to align architecture with user requirements.

Skills

Deep learning

Python

JAX

PyTorch

Reinforcement learning

Research

Education

Masters/PhD in Computer Science or related field

Tools

JAX

PyTorch

Job description

Apple Inc. is seeking an ambitious ML Researcher to shape the future of large-scale foundation models. You will transform models into capable, privacy-forward assistants powering Apple products, tackling instruction following, tool use, and deep reasoning in a fast-growing team.

You will partner with pre-training and product teams, shaping architecture choices and evaluating model performance across practical tasks, while aligning with Apple's privacy and user-centric goals.