Research Engineer, Interpretability Systems 📍 San Francisco Bay Area (On-site) I am partnering with an early-stage AI research lab founded by former frontier-model researchers, focused on alignment and interpretability for large language models. They are building experimental systems and tooling designed to better understand how advanced models reason internally, moving beyond black-box behavior toward mechanistic understanding and controllability. I’m looking for Research Engineers who want to build the experimental infrastructure that makes cutting-edge interpretability research possible. What You’ll Do: • Build custom RL-style environments and experimental testbeds for interpretability research • Develop tooling to probe internal representations, including activation tracing, concept detection, and mechanistic analysis • Implement probes that detect latent concepts such as deception, uncertainty, goals, or hidden objectives • Prototype activation-level steering methods that go beyond prompting or fine-tuning • Work closely with researchers to rapidly iterate from idea → implementation → experiment → result • Help define new benchmarks and measurement frameworks for understanding internal model consistency and robustness • Build tooling that enables entirely new classes of alignment and interpretability experiments Required: • Strong software engineering fundamentals and experience building experimental ML systems • Experience working close to model internals, representations, or post-training systems • Strong Python and deep learning framework experience (PyTorch preferred) • Ability to rapidly prototype and iterate in open-ended research environments • Experience in interpretability, alignment, or ML research preferred • PhD ideal Apply today or contact Ethan - [email protected]