Researcher – Deep Machine Unlearning Methods (DMUMs) (Consultant)
About
We seek a motivated and technically strong researcher to contribute to developing Deep Machine Unlearning Methods (DMUMs). The ideal candidate will be passionate about foundational research, comfortable navigating uncertainty, and capable of producing public-facing outputs.
About the project: to explore and advance the field of machine unlearning – the process of removing specific information from trained AI models. Current unlearning methods face challenges in reliability, resource intensity, and scalability. This project aims to address these limitations by developing robust and scalable Deep Machine Unlearning Methods (DMUMs).
What We’re Looking For:
- 5+ years of background in machine learning, theoretical computer science, or a related field
- Experience working with Deep Machine Unlearning Methods (DMUMs)
- Strong analytical thinking and persistence in resolving complex research questions
- Hands-on experience with Large Language Models (LLMs)
- Clear communication skills, especially in documenting findings and asking clarifying questions when needed
- Prior public contributions (e.g., GitHub repositories, blog posts, academic publications) are a strong plus
- At least Upper-intermediate English
Will be a plus:
- PhD is preferred but not mandatory
- Familiarity with or engagement in relevant research communities is advantageous
Your Responsibilities:
- Develop a deep understanding of the core research problem
- Propose and evaluate approaches to implementing effective DMUMs
- Draft detailed method specifications and research plans
- Investigate unexpected model behaviors and provide data-driven hypotheses
- Document progress regularly, including challenges, assumptions, and milestones
- Demonstrate enthusiasm for potential publication or public dissemination of results