Rosa Lee

• Build scalable, high-quality language datasets and evaluation frameworks that support large language model (LLM) improvement. • Own end-to-end data evaluation and annotation workflows—from problem definition and dataset/evaluation design to execution, quality monitoring, and continuous improvement. • Lead cross-functional initiatives across stakeholders, translating data curation goals into clear workflow design, rubrics, metrics, and deliverables.

• Evaluated state-of-the-art large language models' performance on semantic role processing using methods that draw insights from human language processing research. • Analyzed sentences using language models: extracting word embeddings, probing and obtaining classification accuracies, analyzing attention weights. • Simulated context-driven next-word prediction in adults and children • Developed speech data processing pipelines with Python, R, and speech-to-text APIs. • Published in top-tier SSCI journals and presented at 18 international conferences and workshops. • Conducted psycholinguistic experiments on predictive language processing involving speech and EEG data. • Recruited participants in-person and on online crowdsourcing platforms. • Carried out statistical analyses using Python and R. • Led workshops on computational and experimental methods. • Led a research outreach project in partnership with a language museum in Washington DC. • Led multiple team projects and organized weekly group meetings. • Received merit-based grants and awards based on original research proposals, exceeding $60,000.

• Designed and taught an undergraduate course titled "Prediction in Language Processing." • TAed for 8 linguistics courses and gave lectures on research methods and data analyses • Mentored undergraduate students enrolled in a language science research program

• Co-designed and taught a summer research internship program for Oxford undergraduate students from various academic backgrounds (Computer Science, Philosophy, Psychology, Language Studies). • Led hands-on tutorials on speech data collection using PCIbex and Java, statistical analyses using R, and speech processing pipelines using Google Colab. • Advised on multiple collaborative research projects.

• Collected and analyzed EEG data. • Presented at international conferences. • Wrote sections of original research articles and grant proposals. • Organized lab meetings and trained new lab members on experimental methods and equipment use as the lab manager in the SNU Brain and Humanities Lab.