Anthropic
Research Engineer, Interpretability
$315K–$560K
San Francisco, CA +1 · Full-time · Posted Nov 7, 2025
- Location
- San Francisco, CA
San Francisco, California, United States - Experience
- 5-10+ years building software; the posting also states that required years correlate with internal job-level requirements.
- Education
- Bachelor’s degree or an equivalent combination of education, training, and/or experience; field relevant to the role, demonstrated through coursework, training, or professional experience.
- Visa sponsorship
- Visa sponsorship is offered, but not guaranteed for every role or candidate; if an offer is made, Anthropic will make every reasonable effort to obtain a visa.
About the job
Build and maintain scalable training, inference, and research tooling that enables interpretability research and supports production safety audits of frontier AI models.
What you'll do
- Build specialized training and inference infrastructure, including instrumented model passes, activation extraction, and steering-vector application.
- Profile and optimize systems to address scaling and efficiency bottlenecks, working with infrastructure teams.
- Develop researcher-facing tools and platforms, and help deploy interpretability research in safety audits.
What they ask for
- Highly proficient in at least one programming language and productive with Python.
- Bachelor’s degree or an equivalent combination of education, training, and/or experience.
- Relevant field of study, demonstrated through coursework, training, or professional experience.
- Able to collaborate closely with researchers and translate research needs into engineering solutions.
Nice to have
- Experience optimizing large-scale distributed systems or high-performance LLM workloads.
- Hands-on experience with PyTorch/CUDA on GPUs or JAX/XLA on TPUs.
Pay
$315,000 — $560,000 USD (annual base salary)
Tools
Python, Rust, Go, Java, PyTorch, CUDA, JAX, XLA
AI focus
Build interpretability infrastructure and tooling, and bring interpretability research into production safety audits to support AI safety.
Model-written summary · the original posting is authoritative