Profile
My name is Jin. I studied Electrical and Computer Engineering at RWTH Aachen University (B.Sc., M.Sc.) and did my PhD in Computer Science at Heinrich Heine University Düsseldorf. I am now a postdoctoral researcher in Sikander Hayat's group at Uniklinik RWTH Aachen, working on the EU project AIRIS.
During my PhD I worked on anomaly detection with vision-language models and on making generative models more efficient and safer, with publications at AAAI (oral), ICML and ICLR, in collaboration with groups at the University of Oxford and NUS.
In the group, my focus is kidney disease. I am interested in honest benchmarks for biological foundation models, perturbation prediction on spatial omics from patient tissue, and anomaly detection that separates real biology from technical artifacts. More broadly, I want to understand when we can trust what an AI model tells us about a patient.
Research area
Foundation models and benchmarking, spatial omics, anomaly detection, trustworthy AI for medicine