Join LUMI

Build careful, useful, and trustworthy AI research with us

We welcome inquiries from prospective graduate students, undergraduate research interns, and research collaborators interested in trustworthy language models, reliable agents, multimodal evaluation, and AI safety. We value curiosity, initiative, careful experimentation, clear communication, and respect for collaborators.

yerinhwang@dongguk.edu

Opportunities

Who we are looking for

  • Graduate students

    Prospective graduate students who want to work on trustworthy LLMs, agent reliability, evaluation, or multimodal systems are encouraged to introduce their interests and prior experience.

  • Undergraduate research interns

    Undergraduate students can inquire about research internships. Prior research is helpful but not required; evidence of initiative, solid programming foundations, and willingness to read and test ideas carefully is especially valuable.

  • Collaboration and projects

    We welcome research collaboration, project participation, and proposal discussions with academic and industry partners whose goals align with the research of the lab.

Getting in touch

Writing your first email

There is no required format. Three things are useful to know:

  • a short introduction
  • what you would like to work on, and what draws you to it
  • your CV

Transcripts, code samples, a start date and links to past work are all welcome, though none are expected. If you are unsure whether your background fits, write anyway.

yerinhwang@dongguk.edu

An inquiry does not constitute an admission or appointment offer. Formal university admission requirements and timelines apply separately.

FAQ

Common questions

Do I need a fully formed research topic?

No. A concrete area of interest and evidence that you have explored it are enough for an initial inquiry.

Is prior research required for undergraduate interns?

No, but programming fundamentals, initiative, and willingness to read papers carefully are important.

Can I propose a collaboration?

Yes. Briefly describe the shared problem, expected contribution, and timeline.