Research Scientist, Reasoning & AGI, Google DeepMind

Additional Job Description

We are seeking research scientists to contribute to humanity's critical path to AGI. Our work focuses on reasoning, reinforcement learning (RL), and Deep Think to enhance Gemini and achieve significant breakthroughs for science and humanity. This is a unique opportunity to be at the forefront of AI innovation in Singapore.

The Gemini core team will focus on RL (for pre-training and reasoning), advanced math and coding, and pushing the boundaries of AI. This team collaborates closely with Gemini teams in Mountain View, boasts high talent density, and aims to make a meaningful impact on humanity, science, and Google. Notably, the team was responsible for the Gemini Deep Think breakthrough, which achieved the gold medal standard at the International Math Olympiad (IMO).

Job Responsibilities

• Conduct critical research in RL and reasoning: Propose new RL algorithms, and optimize scaling of RL and inference compute to improve Gemini.
• Serve as Gemini launch captains: Lead model launches and train models to contribute directly to Gemini deployments.
• Innovate and leapfrog: Identify emerging research breakthroughs and build new research agendas based on innovative paradigms.
• Execute on research agenda: Collaborate as a team to implement promising research directions.
• Collaborate with Gemini core teams: Work with teams based primarily in the US and UK.

Qualifications

• PhD in Computer Science or a related STEM field, or equivalent practical experience.
• Strong coding and engineering skills.
• Relevant experience with Large Language Models (LLMs).
• Experience with reinforcement learning and reasoning.

Preferred Qualifications

• Proven impactful research or engineering achievements.
• Exceptional accomplishments in any domain, normalized by circumstances.