Чёткая исследовательская роль в ведущей AI-лаборатории даёт доступ к передовым моделям, понятную научную повестку и возможность публиковаться в значимой области.
Передовой домен
Чёткая научная задача
Современный стек
?Высокий порог входа
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Research Scientist, Mechanistic Interpretability (AI) (mechanistic interpretability and AI safety): Investigating cognitive processes and internal mechanisms in Gemini and other large language models with an accent on self-conceptions, theory of mind, and model representations. Focus on designing empirical experiments, identifying internal vectors that influence outputs, and translating interdisciplinary research into safer and more aligned AGI systems.
Location: London, UK
Company
hirify.global is an AI research lab developing advanced artificial intelligence for scientific discovery, complex global challenges, and safe, aligned AGI.
What you will do
Define and lead a research agenda on mechanistic interpretability in large language models, including self-conceptions, theory of mind, and social media platform-cognition.
Adapt and extend state-of-the-art interpretability techniques for Gemini and other models.
Design and execute experiments to identify internal vectors and mechanisms influencing model outputs and reflective behavior.
Collaborate with technical alignment teams across AI safety and capabilities initiatives.
Publish and communicate findings through peer-reviewed papers, internal technical reports, and presentations for academic and cross-disciplinary audiences.
Requirements
PhD in Computer Science, Mathematics, Physics, or a related technical field, or equivalent research experience.
At least 3 years of experience designing, executing, and analyzing empirical AI research.
Experience using machine learning frameworks such as PyTorch or JAX to run experiments.
Experience with large language model architectures and internal representations.
Experience communicating scientific concepts to technical and non-technical audiences.
Peer-reviewed publication record in mechanistic interpretability, AI safety, or machine learning.
Nice to have
Postdoctoral or equivalent industry research experience in mechanistic interpretability.
Experience or demonstrated interest in cognitive science, philosophy of mind, psychology, or neuroscience.
Experience evaluating selfhood, metacognition, and theory of mind in artificial systems.
Experience translating foundational research into organizational impact, benchmarks, or open-source tools.
Participation in or mentorship of specialized AI safety and interpretability research programs.
Interdisciplinary collaboration bridging technical machine learning with philosophical concepts.
Culture & Benefits
Long-term, ambitious research environment supporting interdisciplinary scientific work.
Global teams with varied learning opportunities and career pathways.
Strong focus on AI safety, ethics, public benefit, and scientific discovery.
Будьте осторожны
Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →