I build foundation models for structured data, particularly tables and graphs, that generalize across datasets and tasks.
I am a founding research scientist at Nums AI and an integrated M.S.–Ph.D. student in Computer Science and Engineering at Seoul National University, advised by Prof. Jaemin Yoo. Previously, I received my B.S. in Electrical Engineering from KAIST.
Minyong Cho*, Minho Jeong*, Dooho Lee*, Jinmo Lee*, Jaemin Yoo* (* equal contribution)
A tabular foundation model combining strong predictive performance with fast inference. Causilo refines cell representations before row compression and is pretrained on approximately 36 million synthetic tables.
Dooho Lee, Myeong Kong, Minho Jeong, Jaemin Yoo
Learns transferable node representations across graphs with heterogeneous feature spaces through Graph View Transformation and a shared pretrained encoder.
Tabular Foundation Model
Use pretrained models for tabular classification and regression through a scikit-learn-compatible interface. Causilo supports numerical and categorical features; the code, model weights, and getting-started guide are linked below.