
Research Engineer / Scientist, Alignment
AI
Anthropic is seeking a Research Engineer/Scientist for Alignment Science in London to design and run machine learning experiments on AI safety, focusing on risks from powerful future systems. The role involves empirical AI research, often collaborating with teams like Interpretability and Fine-Tuning, and requires significant software/ML engineering experience and familiarity with technical AI safety.
What you'll do
- Design and run machine learning experiments to understand and steer AI behavior
- Contribute to exploratory experimental research on AI safety
- Test robustness of safety techniques by training language models to subvert them
What you need
Must-have
PythonMachine LearningReinforcement LearningLLMsKubernetesAI SafetyNLP
Nice-to-have
Research papersShared codebases
Soft skills
CommunicationCollaborationAdaptability
AI SafetyMachine LearningResearch EngineerAlignment ScienceLondonFull-timeOn-sitePythonReinforcement LearningLLMs