This is a fully remote position, open to applicants in United States.

📋 Description

• Create adversarial prompts to rigorously evaluate cutting-edge AI models.

• Detect jailbreaks, unsafe behaviors, hallucinations, and policy shortcomings.

• Assess model resilience in areas such as misinformation, cybersecurity, biosecurity, fraud, political content, and other sensitive sectors.

• Record vulnerabilities and assist in the development of safety benchmarking and red-teaming documentation.

• Partner with AI researchers to enhance model alignment, robustness, and safety measures.

• Engage in projects aimed at training and improving AI systems.

⛳️ Requirements

• A Bachelor's degree or higher in fields such as Computer Science, Cybersecurity, Journalism, Communications, Psychology, Biology, Chemistry, Public Policy, or a related area.

• Over 5 years of professional experience in AI Safety, AI Red Teaming, Trust & Safety, cybersecurity, investigative journalism, life sciences, or a comparable field.

• Excellent analytical reasoning, prompt design capabilities, and written communication skills.

• Proven experience in crafting adversarial prompts or assessing advanced AI systems.

• Preferred: background in AI Red Teaming, Reinforcement Learning from Human Feedback (RLHF), Supervised Fine-Tuning (SFT), AI Alignment, or Trust & Safety.

• Preferred: knowledge of jailbreak testing, prompt engineering, or adversarial evaluation techniques.

• Preferred: specialization in cybersecurity, biosecurity, political content, misinformation, or scientific safety.

• Candidates must not require H1-B or STEM OPT support; Mercor is unable to assist H1-B or STEM OPT applicants.