Research Engineer, Frontier Safety Risk Assessment, Deepmind

Google Google · Big Tech · San Francisco, CA +3

Research Engineer focused on assessing risks from frontier AI models, identifying risk pathways, and developing methods for measuring risk. Involves implementing research findings and building technical solutions using Python, with experience in deep learning and large foundation models.

What you'd actually do

  1. Drive novel research to measure, assess, and prioritize risks from frontier AI models.
  2. Identify and analyze emerging risk pathways in areas including loss of control, Machine Learning (ML) Research and Development (R&D), cybersecurity, and harmful manipulation.
  3. Design and develop new methods for measuring pre-mitigation and post-mitigation risk, incorporating forecasting and scenario planning.
  4. Implement research findings and build technical solutions using Python, interacting with various codebases.
  5. Prioritize research and engineering efforts within pragmatic constraints of compute and time to maximize value of information.

Skills

Required

  • deep learning
  • large foundation models
  • AI experiments
  • AI systems
  • Python
  • assessing AI model risks
  • mitigating AI model risks
  • technical stakeholders

Nice to have

  • Ph.D. in Computer Science, Machine Learning, or a related technical field
  • frontier risk assessment
  • mitigations
  • AI safety
  • alignment
  • applied research projects
  • LLM training
  • LLM inference
  • publications at NeurIPS, ICLR, ICML, RL/DL workshops, EMNLP, AAAI, or UAI

What the JD emphasized

  • frontier AI models
  • assessing risks
  • risk pathways
  • AI model risks

Other signals

  • frontier AI models
  • risk assessment
  • AI safety
  • alignment