Job Description

Minimum qualifications

• PhD degree in Computer Science, a related field, or equivalent practical experience.
• 2 years of experience with Large Language Model internals or generalization.

Preferred qualifications

• Prior publication (NeurIPS/ICML/ICLR) or equivalent milestone, in the area of information propagation, generalisation, or mechanistic interpretability of intelligent systems.
• Experience with leading research projects.
• Strong experimental taste with good judgment regarding baselines, ablations, and what is worth testing.
• Interest in Safety and Security of Generative AI systems.
• A generalist demonstrating adaptability and flexibility across several research projects.
• Demonstrated ability to work collaboratively as well as independently.

About the job

As a part of the Frontier Computation team at DeepMind, you will focus on the intricate relationship neural networks have with (classical) computation, which entails both improving the algorithmic alignment properties of frontier AI systems, and using frontier AI systems to improve the substrate of computation.

In this friendly, tight-knit, focused and collaborative team, you will explore a broad mixture of topics on the frontier of machine learning, spanning long-term research moonshots as well as direct deployments into DeepMind's AI offerings.

As a Research Scientist in our team, you would ideally make foundational contributions across a wide range of these projects, as and when appropriate given your expertise. You will also demonstrate a continuous learning mindset, and expand the range of topics you can meaningfully contribute to at GDM more broadly. In this role specifically, you will place emphasis on our growing engagements with Security and Safety research within generative AI systems.Artificial intelligence will be one of humanity’s most transformative inventions. At Google DeepMind, we are a pioneering AI lab with exceptional interdisciplinary teams focused on advancing AI development to solve complex global challenges and accelerate high-quality product innovation for billions of users. We use our technologies for widespread public benefit and scientific discovery, ensuring safety and ethics are always our highest priority.

We are pushing the boundaries across multiple domains. Our global teams offer varied learning opportunities and career pathways for those driven to achieve exceptional results through collective effort.

Responsibilities

• Lead long-term "moonshot" projects and direct product deployments, probing model internals, mechanistic interpretability, and information propagation with high personal initiative, independently design, implement, and execute rigorous empirical experiments, applying strict criteria for baselines and ablation studies to systematically validate research directions.
• Take ownership of research engagements in AI security, developing prompt defenses and alignment strategies to secure our frontier systems.
• Translate complex findings into clear, structured results and drive be involved into close collaboration with internal and external collaborators.