Requirements

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

Nice to haves

• Prior publication (NeurIPS/ICML/ICLR) or equivalent milestone in information propagation, generalisation, or mechanistic interpretability of intelligent systems.
• Experience with leading research projects.
• Strong experimental taste with good judgment on baselines, ablations, and meaningful tests.
• Interest in Safety and Security of Generative AI systems.
• A generalist adaptable across several research projects.
• Demonstrated ability to work collaboratively as well as independently.

What you'll be doing

• Lead long-term "moonshot" projects and direct product deployments focusing on model internals, mechanistic interpretability, and information propagation.
• Independently design, implement, and execute rigorous empirical experiments with strict criteria for baselines and ablation studies.
• Take ownership of AI security research including prompt defenses and alignment strategies to secure frontier systems.
• Translate complex findings into clear, structured results and collaborate closely with internal and external partners.