Rahul Thapa

I am currently working with Dr. James Zou. Previously, I worked as a research engineer in Dr. Nigam Shah’s lab. My current research interest revolves around training multi-modal foundation models and evaluating them within the realm of healthcare. I am pursuing my Ph.D. with the generous support of the Knight-Hennessy Scholarship program at Stanford University.

I graduated summa cum laude from Villanova University in 2021 with a degree in Computer Science. During my time at Villanova, I worked very closely with Dr. Xun Jiao on multiple research projects surrounding approximate computing and hyperdimensional computing.

News

Jun 10, 2024	EHRSHOT, a dataset of 6,739 deidentified longitudinal EHRs for few-shot eval of foundation models, is finally out. Check us out!
Jun 3, 2024	Dragonfly out on arXiv. We proposed a new architecture for vision-language models, and showed promising results on biomedical benchmarks, even beating Med-gemini on multiple tasks.
May 1, 2024	SleepFM accepted to ICML 2024. We built the first foundation model for sleep analysis. See you in Vienna!
Apr 1, 2024	Started my research internship at Together AI. Working to build better open-source multimodal models.
Sep 26, 2023	My paper on Machine Learning Differentiation of Autism Spectrum Sub-Classifications accepted at Journal of Autism and Developmental Disorders
Sep 26, 2023	Started PhD in Biomedical Data Science at Stanford University
Sep 22, 2023	EHRSHOT accepted at NeurIPS 2023 Datasets and Benchmark Track
May 9, 2023	Selected as a 2023 Knight-Hennessy Scholar at Stanford University
Aug 14, 2022	Joined Stanford Center for Biomedical Informatics Research as a Data Scientist, working closely with Dr. Nigam Shah
Feb 15, 2022	Left Dascena to join Forta as a Senior Data Scienct

Selected Publications

NeurlPS

Ehrshot: An ehr benchmark for few-shot evaluation of foundation models

Michael Wornow*, Rahul Thapa*, Ethan Steinberg, and 2 more authors

arXiv preprint arXiv:2307.02028, 2023

Bib Code Website

@article{wornow2023ehrshot,
  title = {Ehrshot: An ehr benchmark for few-shot evaluation of foundation models},
  author = {Wornow*, Michael and Thapa*, Rahul and Steinberg, Ethan and Fries, Jason and Shah, Nigam},
  journal = {arXiv preprint arXiv:2307.02028},
  year = {2023},
  url = {https://som-shahlab.github.io/ehrshot-website/},
  paper = {https://arxiv.org/abs/2307.02028},
  dimensions = {true},
}

Springer

Machine Learning Differentiation of Autism Spectrum Sub-Classifications

R Thapa, A Garikipati, M Ciobanu, and 7 more authors

Journal of Autism and Developmental Disorders, 2023

Bib Website

@article{thapa2023machine,
  title = {Machine Learning Differentiation of Autism Spectrum Sub-Classifications},
  author = {Thapa, R and Garikipati, A and Ciobanu, M and Singh, NP and Browning, E and DeCurzio, J and Barnes, G and Dinenno, FA and Mao, Q and Das, R},
  journal = {Journal of Autism and Developmental Disorders},
  pages = {1--16},
  year = {2023},
  publisher = {Springer},
}

Elsevier

Assessing the effects of data drift on the performance of machine learning models used in clinical sepsis prediction

Keyvan Rahmani*, Rahul Thapa*, Peiling Tsou, and 4 more authors

International Journal of Medical Informatics, 2023

Bib Website

@article{rahmani2023assessing,
  title = {Assessing the effects of data drift on the performance of machine learning models used in clinical sepsis prediction},
  author = {Rahmani*, Keyvan and Thapa*, Rahul and Tsou, Peiling and Chetty, Satish Casie and Barnes, Gina and Lam, Carson and Tso, Chak Foon},
  journal = {International Journal of Medical Informatics},
  volume = {173},
  pages = {104930},
  year = {2023},
  publisher = {Elsevier},
}

JMIR

Multitask Learning With Recurrent Neural Networks for Acute Respiratory Distress Syndrome Prediction Using Only Electronic Health Record Data: Model Development and Validation Study

Carson Lam*, Rahul Thapa*, Jenish Maharjan, and 5 more authors

JMIR Medical Informatics, 2022

Bib Website

@article{lam2022multitask,
  title = {Multitask Learning With Recurrent Neural Networks for Acute Respiratory Distress Syndrome Prediction Using Only Electronic Health Record Data: Model Development and Validation Study},
  author = {Lam*, Carson and Thapa*, Rahul and Maharjan, Jenish and Rahmani, Keyvan and Tso, Chak Foon and Singh, Navan Preet and Casie Chetty, Satish and Mao, Qingqing},
  journal = {JMIR Medical Informatics},
  volume = {10},
  number = {6},
  pages = {e36202},
  year = {2022},
  publisher = {JMIR Publications Toronto, Canada},
}

JMIR

Predicting falls in long-term care facilities: machine learning study

Rahul Thapa, Anurag Garikipati, Sepideh Shokouhi, and 7 more authors

JMIR aging, 2022

Bib Website

@article{thapa2022predicting,
  title = {Predicting falls in long-term care facilities: machine learning study},
  author = {Thapa, Rahul and Garikipati, Anurag and Shokouhi, Sepideh and Hurtado, Myrna and Barnes, Gina and Hoffman, Jana and Calvert, Jacob and Katzmann, Lynne and Mao, Qingqing and Das, Ritankar},
  journal = {JMIR aging},
  volume = {5},
  number = {2},
  pages = {e35373},
  year = {2022},
  publisher = {JMIR Publications Toronto, Canada},
}

Pancreatology

Early prediction of severe acute pancreatitis using machine learning

Rahul Thapa, Zohora Iqbal, Anurag Garikipati, and 4 more authors

Pancreatology, 2022

Bib Website

@article{thapa2022early,
  title = {Early prediction of severe acute pancreatitis using machine learning},
  author = {Thapa, Rahul and Iqbal, Zohora and Garikipati, Anurag and Siefkas, Anna and Hoffman, Jana and Mao, Qingqing and Das, Ritankar},
  journal = {Pancreatology},
  volume = {22},
  number = {1},
  pages = {43--50},
  year = {2022},
  publisher = {Elsevier},
}

IEEE

Spamhd: Memory-efficient text spam detection using brain-inspired hyperdimensional computing

Rahul Thapa, Bikal Lamichhane, Dongning Ma, and 1 more author

In 2021 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2021

Bib Website

@inproceedings{thapa2021spamhd,
  title = {Spamhd: Memory-efficient text spam detection using brain-inspired hyperdimensional computing},
  author = {Thapa, Rahul and Lamichhane, Bikal and Ma, Dongning and Jiao, Xun},
  booktitle = {2021 IEEE Computer Society Annual Symposium on VLSI (ISVLSI)},
  pages = {84--89},
  year = {2021},
  organization = {IEEE},
}

IEEE

Hdxplore: Automated blackbox testing of brain-inspired hyperdimensional computing

Rahul Thapa, Dongning Ma, and Xun Jiao

In 2021 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2021

Bib Website

@inproceedings{thapa2021hdxplore,
  title = {Hdxplore: Automated blackbox testing of brain-inspired hyperdimensional computing},
  author = {Thapa, Rahul and Ma, Dongning and Jiao, Xun},
  booktitle = {2021 IEEE Computer Society Annual Symposium on VLSI (ISVLSI)},
  pages = {90--95},
  year = {2021},
  organization = {IEEE},
}