Research Engineer - Oversight Foundations
Transluce · San Francisco, United States
Research instituteonsite$250k–$500k
At a glance
Highlights
- Build oversight foundation models that formalize and test AI system behavior.
- Develop assistants that predict and detect unexpected behaviors in AI systems.
- Scale training and inference pipelines to support up to 1T-scale models.
Ideal candidate
A curious, low-ego scientist-engineer who loves understanding how ML systems work and inventing ways to probe them. You bring strong experimentation instincts and enjoy collaborating in person to solve frontier oversight problems.
Requirements
- Experience with fine-tuning language models, designing architectures, and creating evaluations
- Strong experimental design with epistemic self-awareness and transparency
- Generativeness in coming up with original ideas to unblock progress
- Curiosity to understand how ML systems work
- Strong programming ability balancing prototyping speed and maintainability
- Strong communication skills, low ego, openness to feedback
- remote policy
- In-person
- visa sponsorship
- Open to sponsoring international visas
Lab profile
About Transluce
Non-profit research lab building public tech stack for AI evaluation and oversight.
What they build
- Docent agent-analysis platform
- Behavioral evaluations of frontier AI systems
- AI-backed tools for understanding AI
Known for
- Evaluation of emergent high-stakes AI behaviors
- Scalable AI oversight and debugging
- Work shaping AI development and governance
Key facts
- Independent non-profit research lab
Posted · Last seen