LogoFresh Jobs Ahead
Mercor
Mercor
51 – 200 Darbuotojai
Turimame Mercor profilyje nėra pakankamai patikrintos informacijos apie įmonės produktus, paslaugas, klientus, veiklos sritį ar įdarbinimo kryptį, kad būtų galima tiksliai apžvelgti įmonę. Prieš aprašant įmonę ar jos teikiamas galimybes, reikia daugiau šaltinių medžiagos.
LinkedIn
Darbai įmonėje

AI Safety Red Teamer

Mercor seeks an AI Safety Red Teamer to probe frontier models with adversarial prompts. The work focuses on finding weaknesses and helping AI lab partners improve model alignment, robustness, and safety.

USJungtinės Valstijos
70 $ – 84 $ / valanda
NuotolinisSutartisVidutinis lygisVyresnysis specialistasPrivaloma Kalba: anglųDirbtinis intelektas
Kibernetinis saugumas

Aprašymas

  • Create adversarial prompts to probe frontier AI models.
  • Find jailbreaks, unsafe behavior, hallucinations, and policy failures.
  • Assess model robustness in sensitive areas including misinformation, cybersecurity, biosecurity, fraud, and political content.
  • Record vulnerabilities and help prepare safety benchmarks and red-teaming reports.
  • Work with AI researchers to strengthen model alignment, robustness, and safety.
  • Contribute to projects that train and improve AI systems.

Reikalavimai

  • Bachelor’s degree or higher in computer science, cybersecurity, journalism, communications, psychology, biology, chemistry, public policy, or a related field.
  • At least five years of professional experience in AI safety, AI red teaming, trust and safety, cybersecurity, investigative journalism, life sciences, or a related area.
  • Strong analytical reasoning, prompt design, and written communication abilities.
  • Experience creating adversarial prompts or assessing frontier AI systems.
  • Experience with AI red teaming, RLHF, SFT, AI alignment, or trust and safety is preferred.
  • Familiarity with jailbreak testing, prompt engineering, or adversarial evaluation methods is preferred.
  • Expertise in nuanced areas such as cybersecurity, biosecurity, political content, misinformation, or scientific safety is preferred.
  • Ability to work as an independent contractor.
  • H-1B and STEM OPT candidates are not supported.

Privalumai

  • Work remotely from anywhere in the United States.
  • Set your own work schedule.
  • Receive weekly payments through Stripe or Wise.
  • Projects may be extended, shortened, or ended early based on project needs and performance.
  • Work with leading AI researchers and safety teams.
  • Help shape the safety of frontier AI systems.
  • Reasonable accommodations are available upon request.
  • Earn referral payments of up to $340 per successful referral; limits apply.

Susiję darbai