Research Scientist (Reinforcement Learning)

TL;DR

Research Scientist (Reinforcement Learning): Conducting end-to-end research experiments and developing advanced algorithms for reinforcement learning systems with an accent on hypothesis testing, ablation studies, and rigorous evaluation. Focus on building scalable research infrastructure and communicating findings through high-impact publications.

Company

DeepMind is a world-leading AI research organization dedicated to solving intelligence to advance science and benefit humanity.

What you will do

• Formulate and test research hypotheses through end-to-end experimentation.
• Implement and iterate on complex algorithms within research codebases.
• Design and execute rigorous ablation studies and model evaluations.
• Analyze experimental results to derive actionable insights.
• Build and improve the underlying research infrastructure.
• Communicate research findings through technical write-ups and academic publications.

Requirements

• Strong background in Reinforcement Learning and Deep Learning.
• Proficiency in research-oriented programming using JAX, Flax, or PyTorch.
• Experience in designing and conducting complex machine learning experiments.
• Ability to perform deep data analysis and interpret experimental outcomes.
• Strong communication skills for documenting and presenting research.