Skip to main content

Research Scientist, Gemini Agent Safety Post Training, DeepMind

Summary by AIEngineer.careers

Research Scientist, Gemini Agent Safety Post Training, DeepMind is a AI Safety Engineer role (full-time). with Google. in MOUNTAIN VIEW, US. Compensation shown: $207K–$300K. Imported listing (source: movojobs.com). Apply on the employer's site (movojobs.com).

Imported listing (source: movojobs.com) · Apply on movojobs.com

How our listings work

MOUNTAIN VIEW · US
$207K–$300K
Full-time
MID
Apply Now

More AI jobs in United States

The sections below reproduce the third-party job description for reference. AIEngineer.careers does not write or control this text.

Imported job description

Sourced from movojobs.com

Join Google DeepMind's Gemini Safety team as a Research Scientist focused on agent safety post-training. You will develop algorithmic solutions and alignment techniques to identify and close safety gaps across Gemini and GenAI models, ensuring outputs are aligned and production-ready. The role combines LLM post-training research with hands-on engineering to ship safety improvements at massive scale.

Responsibilities

  • Design and maintain protocols to identify safety gaps in model behavior across Gemini and GenAI models
  • Feed safety insights back into the post-training process to improve in-model and out-of-model protection
  • Apply innovative modeling and alignment techniques to solve real-world safety issues under tight launch timelines
  • Advance Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) techniques at massive scale
  • Partner with post-training, evaluation, and product teams to align incentives and close safety gaps

Requirements

  • PhD in Computer Science, a related field, or equivalent practical experience
  • 3 years of experience initiating, scoping, and executing research agendas or leading academic projects
  • One or more peer-reviewed scientific publications (e.g., NeurIPS, ICML, ICLR, CVPR, ACL, CoRL) or public preprints (arXiv)
  • LLM post-training expertise to ensure outputs are aligned and production-ready

Nice to Have

  • Experience in AI safety and model alignment
  • Experience developing AI agents, safety-critical systems, or Machine Learning Infrastructure
  • Expertise in reward modeling, RL for LLMs, instruction tuning, and long-range RL
  • Track record of publications at AI/ML venues (e.g., NeurIPS, ICLR, ICML, EMNLP, AAAI, UAI)
  • Ability to drive research concepts to product realization with strong engineering abilities

Benefits

Equity / stock optionsbonusHealth insuranceretirement plan

Tech Stack

AI AgentsGeminiInstruction TuningLLMsMachine Learning InfrastructureModel AlignmentReinforcement LearningReward ModelingSupervised Fine-Tuning
Google

Google

Mountain View · US · 297958+ employees

Google is a global technology company specializing in internet-related services and products, including search, cloud computing, advertising technologies, and software. It is a subsidiary of Alphabet Inc.

Weekly job alert

Get a weekly email when new AI engineering jobs match your filters. Free — confirm your address to start.

Related Jobs