Artificial Intelligence Researcher — United States
Зарплата не указана
ОткликнутьсяКомпания: ••••••••
Локация: United States
Формат: не указан
Занятость: Полная занятость
Грейд: не определён
AI Alignment & Red-Teaming Researcher
Experience: 0–5 Years
Employment Type: Full-Time
Work Mode: [On-site / Hybrid / Remote]
Job Summary
We are seeking an AI Alignment & Red-Teaming Researcher to evaluate the safety, reliability, and behavior of AI systems, particularly Large Language Models (LLMs). The ideal candidate will investigate potential vulnerabilities, identify harmful or unintended model behaviors, conduct adversarial testing, and contribute to improving AI alignment, robustness, and responsible AI development.
Key Responsibilities
• Conduct red-teaming exercises to identify vulnerabilities and weaknesses in AI models and applications.
• Evaluate LLM responses for harmful, biased, misleading, or unsafe outputs.
• Design and execute adversarial prompts and testing scenarios to assess model behavior.
• Identify risks such as prompt injection, jailbreaks, data leakage, and hallucinations.
• Assess AI systems for safety, fairness, reliability, and policy compliance.
• Develop test cases, evaluation frameworks, and benchmarks for AI safety testing.
• Analyze model outputs and document vulnerabilities, findings, and recommendations.
• Collaborate with AI engineers, researchers, and product teams to improve model safety.
• Investigate emerging AI alignment techniques and responsible AI practices.
• Contribute to research documentation, technical reports, and evaluation methodologies.
Required Skills
• Understanding of Artificial Intelligence, Machine Learning, and Large Language Models.
• Knowledge of prompt engineering and adversarial testing techniques.
• Familiarity with LLM vulnerabilities, including prompt injection and jailbreak attacks.
• Understanding of AI safety, responsible AI, and model evaluation principles.
• Basic programming knowledge in Python.
• Familiarity with NLP concepts and generative AI applications.
• Strong analytical, critical-thinking, and research skills.
• Ability to design structured experiments and interpret evaluation results.
• Strong technical writing and documentation skills.
Preferred Qualifications
• Bachelor's or Master's degree in Computer Science, Artificial Intelligence, Machine Learning, Data Science, Mathematics, or a related field.
• Hands-on experience evaluating LLMs or conducting AI red-teaming projects.
• Familiarity with frameworks and tools such as Hugging Face, PyTorch, TensorFlow, or LangChain.
• Exposure to AI evaluation and security tools such as Garak, PyRIT, or promptfoo.
• Understanding of adversarial machine learning, model robustness, and AI governance.
• Research experience, academic publications, or independent AI safety projects.