Aiml - ML Engineer, Responsible AI

Apple Apple · Big Tech · Seattle, WA +2 · Machine Learning and AI

ML Engineer focused on Responsible AI, developing models, tools, and metrics for assessing and evaluating the safety, robustness, and uncertainty of generative models (vision and language). This includes interpreting model failures, building human annotation and red teaming pipelines, and prototyping/implementing/evaluating new ML models for red teaming LLMs.

What you'd actually do

  1. Develop models, tools, metrics, and datasets for assessing and evaluating the safety of generative models over the model deployment lifecycle
  2. Develop methods, models, and tools to interpret and explain failures in language and diffusion models
  3. Build and maintain human annotation and red teaming pipelines to assess quality and risk of various Apple products
  4. Prototype, implement, and evaluate new ML models and algorithms for red teaming LLMs

Skills

Required

  • Strong engineering skills
  • writing production-quality code in Python, Swift or other programming languages
  • generative models
  • natural language processing
  • LLMs
  • diffusion models
  • failure analysis
  • quality engineering
  • robustness analysis for AI/ML based features
  • working with crowd-based annotations and human evaluations
  • explainability and interpretation of AI/ML models
  • Work with highly-sensitive content with exposure to offensive and controversial content

Nice to have

  • BS, MS or PhD in Computer Science, Machine Learning, or related fields or an equivalent qualification acquired through other avenues
  • Proven track record of contributing to diverse teams in a collaborative environment

What the JD emphasized

  • safety
  • robustness
  • uncertainty
  • failure analysis
  • quality engineering
  • robustness analysis
  • explainability
  • interpretation of AI/ML models
  • offensive and controversial content

Other signals

  • Responsible AI
  • generative AI
  • evaluating safety
  • interpret and explain failures
  • red teaming LLMs