[Expression of Interest] Research Engineer / Scientist, Alignment - London
This role involves conducting experimental research in AI safety and alignment, focusing on risks from advanced AI systems. You'll design and run machine learning experiments—such as testing safety robustness, developing control methods, and stress-testing alignment—to improve the reliability and steerability of AI. The work is highly collaborative, involving contributions to research papers and close coordination with teams like Interpretability and Fine-Tuning.