About

I research machine learning methods, lately focusing on evaluation of reasoning capabilities of large language models, synthetic datasets, reproducibility, and applications in medicine.

I received my MA and am continuing on with the PhD at the Computer Lab at the University of Cambridge, advised by Prof Ferenc Huszár. In the past, I spent some time at the University of Oxford for the MSc and Google for three internships.

For academic years 23/24 and 24/25 I am visiting Prof Kilian Q. Weinberger’s lab at Cornell.

Publications and preprints

PhantomWiki: on-demand datasets for reasoning and retrieval evaluation. Albert Gong*, Kamilė Stankevičiūtė*, Chao Wan*, Anmol Kabra, Raphael Thesmar, Johann Lee, Julius Klenke, Carla P. Gomes, Kilian Q. Weinberger. International Conference on Machine Learning (ICML) 2025. Oral presentation at the ICML 2025 Workshop on Long-Context Foundation Models.
[pdf] [arxiv] [code]

Bridging the worlds of pharmacometrics and machine learning. Kamilė Stankevičiūtė*, Jean-Baptiste Woillard*, Richard W. Peck, Pierre Marquet, Mihaela van der Schaar. Clinical Pharmacokinetics (2023). [pdf]

Conformal time-series forecasting. Kamilė Stankevičiūtė, Ahmed M. Alaa, Mihaela van der Schaar. Neural Information Processing Systems (NeurIPS) (2021). Lithuania’s best early-career AI publication of 2021. Best short presentation at the 10th Meeting of Early Career Mathematicians of Lithuania. [pdf] [code] [video]

Population graph GNNs for brain age prediction. Kamilė Stankevičiūtė, Tiago Azevedo, Alexander Campbell, Richard A. I. Bethlehem, Pietro Liò. ICML Workshop on Graph Representation Learning and Beyond (2020). M2L outstanding poster award. [pdf] [code] [video]

Software

MEDS Decentralized, Extensible Validation (MEDS-DEV) Benchmark. MEDS-DEV Working Group. Hands-On Tutorial at KDD 2025. Demo at ML4H 2024. Invited talk at the 13th Meeting of Early Career Mathematicians of Lithuania. [spec sheet] [docs] [code]