Research Scientist, Gemini Horizon: AI Systems & RL

Google LLC; Mountain View (CA); On-site; USD 174,000 - 252,000; Full time

Job summary

Google DeepMind is seeking a Research Scientist for Gemini Horizon to drive large-scale testing and rapid deployment of innovative ideas. You will work on architectures spanning machine learning, data mining, NLP, and systems design, with emphasis on real-world impact and collaboration across global teams.

Role involves publishing findings, collaborating with academia and partner programs, and contributing to responsible AI and safety practices, with competitive compensation and equity as part

Qualifications

• 2 years of experience with machine learning algorithms and tools, or applied ML (e.g., LLMs, generative AI, NLP, recommendations).
• 2 years of experience with Generative AI, AI algorithms, and system design.
• 2 years of industry experience in applied machine learning/machine learning research.
• 2 years of experience with developing machine learning infrastructure, distributed systems or networks.

Responsibilities

• Deliver step-change improvements to Gemini through rigorous and ambitious research.
• Improve the models ability to manage its own context and memory natively, instead of relying on the code wrapped around the model (the harness).
• Make Gemini better at subagent orchestration, which is basically memory management through isolation.
• Advance complex long horizon task execution such as for RSI and computer use where we combine low latency asynchronous execution and long horizon reasoning.
• Build novel infrastructure, gaining a deep understanding of training dynamics, and developing novel RL algorithms.

Skills

Machine learning

Generative AI

Python

C++

Education

M.S. or PhD in Computer Science/ML

Tools

JAX

PyTorch

Job description

Google DeepMind is seeking a Research Scientist for Gemini Horizon to drive large-scale testing and rapid deployment of innovative ideas. You will work on architectures spanning machine learning, data mining, NLP, and systems design, with emphasis on real-world impact and collaboration across global teams.

Role involves publishing findings, collaborating with academia and partner programs, and contributing to responsible AI and safety practices, with competitive compensation and equity as part