Staff Research Scientist, STEM

Tech Stack

Responsibilities

• Identify high-impact gaps in existing benchmark and evaluation literature and design novel benchmarks in and across STEM fields.
• Develop methods for generating high-quality synthetic STEM training data and study downstream performance impacts.
• Study hallucination, uncertainty, calibration, and epistemic failure in technical domains to improve factual reliability.
• Research and evaluate long-horizon scientific agents and workflows capable of performing extended scientific work.
• Propose and launch ambitious new research programs in areas such as reasoning, model evaluation, and AI-for-science.

Soft Skills

Benefits

• Equity

Culture

Requirements