Research Scientist, Gemini Agent Safety Post Training, DeepMind

What you'll do

• Create safety protocols to identify gaps in model behavior and feed findings into post-training.
• Apply modeling and alignment techniques, including SFT and RL, to improve safety at scale.
• Coordinate with post-training, evaluation, and product teams to close safety gaps and guide launch decisions.

What they ask for

• PhD in Computer Science, a related field, or equivalent practical experience.
• 1 year experience with Generative AI, Large Language Models, natural language processing, or Agent-based systems.

Nice to have

• Experience in AI safety and model alignment; AI agents, safety-critical systems, or Machine Learning Infrastructure.
• Expertise in reward modeling, RL for LLMs, instruction tuning, and long-range RL; product realization, engineering, experimental work, or AI/ML publications.

Pay

$174000 - $252000 USD (base salary)

Tools

Gemini, GenAI, Supervised Fine-Tuning (SFT), Reinforcement Learning (RL), Large Language Models, natural language processing, Machine Learning Infrastructure

AI focus

AI safety and alignment post-training, model behavior evaluation, safety mitigation, and large-scale SFT and RL optimization.