LLM Security Researcher

The LLM Security Researcher - Red-Teamer role at Crossing Hurdles involves developing adversarial multi-turn conversations and task-based scenarios for AI model evaluation. Your responsibilities include crafting clear evaluation rubrics, testing frontier LLMs, validating outputs, and documenting model strengths and failure modes. You will work autonomously, interpreting complex specifications, and collaborating with team leads to ensure alignment with project requirements. Ideal candidates have exceptional English skills, deep AI expertise (RLHF, prompt engineering, model evaluation), and experience designing evaluation items. This is a contract position offering flexible hours (10-40/week) with a remote work arrangement.