Description
Anthropic is hiring an AI Infrastructure Engineer for its Interpretability team to build and maintain specialized inference and training infrastructure, including instrumented forward and backward passes, activation extraction, steering-vector application, profiling, optimization, and research tooling. The role works across model internals, accelerator-level optimization, and user-facing tools, supports production safety audits, and requires 5–10+ years of software-building experience, strong Python proficiency, and a relevant bachelor’s degree or equivalent education and experience. The role is based in the San Francisco office with a case-by-case remote-work option and an annual salary of $315,000–$560,000 USD.
