Kyungeun is an AI safety researcher with a PhD in Physics from Columbia and postdoctoral research at Yale. She played key leadership roles in international collaborations in dark matter and neutrino physics, leading contributions to over 40 peer-reviewed publications with more than 8,800 citations, and has served as a reviewer for physics journals. Her work spanned experimental design, detector construction, data analysis, and publication, with hands-on experience at laboratories including the Kamioka Observatory in Japan and Gran Sasso National Laboratory in Italy. Through her work on the CUORE and XENON experiments, she developed expertise in cross-cultural scientific collaboration and systematic experimental methodology, and took on leadership roles including data production lead and vetting board member for CUORE.
She currently works as an independent AI safety researcher, supported by BlueDot Impact grants, studying what safety interventions actually change inside language models, from red-teaming their behavior to analyzing their internals. This grew out of her fellowship at ERA:AI Cambridge, where she analyzed how machine unlearning modifies model weights and released a model variant from that work on HuggingFace. Before AI safety, she worked in industry ML, developing production systems at NBCUniversal for major events like the Olympics and Super Bowl, and at CloudTrucks where she built end-to-end ML product pipelines including personalized recommendation engines. See the Projects tab for more details.
Beyond her technical work, she is an avid learner who engages with podcasts, audiobooks and videos covering philosophy, big history, art, fashion, classical music, and literature. She maintains an active lifestyle through regular dance classes (for over a decade), barre, pilates, and yoga, complemented by frequent walks and monthly museum visits. She enjoys traveling, with extensive experiences across Europe, the Americas, and Africa. She likes tango and bossa nova, as well as K-pop music, and recently enjoys reading sci-fi. She preserves her attention by avoiding social media, choosing instead to nurture meaningful friendships through weekly or monthly conversations.
The Ella Project, May 2018
Profiled as a Senior Lead Data Scientist discussing barriers and opportunities for women in STEM careers, sharing insights on overcoming stereotypes and building confidence in technical fields.
Insight Data Science, March 2018
Participated in panel discussion at Columbia University offering career transition advice to students, sharing experiences from academia to industry data science.
New York Times, April 2011
Featured in coverage of dark matter research during graduate studies at Columbia University.
Selected presentations at major research institutions