Hi, nice to meet you!
I'm a PhD student in the Scalable Trustworthy AI group at the University of Tübingen advised by Seong Joon Oh as part of IMPRS-IS's doctoral program. With my research I would like to contribute to researching how technology can benefit humans and society, in particular the field of responsible and trustworthy AI. I believe that AI offers a great potential to improve our lives and contribute to solving issues but technology must not cause or facilitate harms. Before starting the PhD, I absolved my Master's in Interaction Technology at the University of Twente and my Bachelor's in Computer Science and Management at the Nordakademie. I did my Bachelor studies in a dual study format in cooperation with Airbus, where I also worked for about 2.5 years as an IT project manager for CAD software in waterfall and agile projects. Currently, I am interested in data attribution as explanations. How do models learn from data and is it possible to root a learned behavior in the training data? How can data attribution best be utilized to empower humans to build and interact with trustworthy models? Feel free to reach out if you would like to get in touch:
Updates
- 26 Sep 24: New preprint alert! This one has been a year+ long work on understanding user needs for training data attribution. Read it here.
- 26 Sep 24: I'm excited to be a PhD representative for the Tübingen AI Center this upcoming year.
- 14 Jul 24: I'll be co-organising the ATTRIB workshop with Tolga Bolukbasi, Logan Engstrom, Andrew Ilyas, Sadhika Malladi, and Sam Park at NeurIPS'24!
- 10 Jul 24: Evgenii's thesis work was accepted to COLM.
- 25 Apr 24: Evgenii graduated last month! It was a pleasure working with him. Read about his Master thesis research on language model behaviour under knowledge conflicts in our new preprint.
- 27 Oct 23: Our interview study about ML practioner perspectives was accepted to the XAIA workshop at NeurIPS! Read it here.
- 21 Sep 23: My first PhD paper was accepted to NeurIPS! If you're interested in the reliability of training data attribution, give it a read here.
Publications
* equal contributionTowards User-Focused Research in Training Data Attribution for Human-Centered Explainable AI Elisa Nguyen, Johannes Bertram, Evgenii Kortukov, Jean Y. Song, Seong Joon Oh arxiv 2024 [paper] |
Studying Large Language Model Behaviors Under Realistic Knowledge Conflicts Evgenii Kortukov, Alexander Rubinstein, Elisa Nguyen, Seong Joon Oh COLM 2024 [paper][code] |
Exploring Practitioner Perspectives On Training Data Attribution Explanations Elisa Nguyen, Evgenii Kortukov, Jean Y. Song, Seong Joon Oh XAI in Action Workshop at NeurIPS 2023 [paper][poster] |
A Bayesian Approach To Analysing Training Data Attribution In Deep Learning Elisa Nguyen, Minjoon Seo, Seong Joon Oh NeurIPS 2023 [paper][code][poster][conf page incl. video+slides] |
Feature Attribution Explanations for Spiking Neural Networks Elisa Nguyen, Meike Nauta, Gwenn Englebienne, Christin Seifert CogMI 2023 [paper][code] |
Trustworthy Machine Learning Bálint Mucsányi, Michael Kirchhof, Elisa Nguyen, Alexander Rubinstein, Seong Joon Oh arxiv 2023 [book][website] |
From Anecdotal Evidence to Quantitative Evaluation Methods: A Systematic Review on Evaluating Explainable AI Meike Nauta, Jan Trienes, Shreyasi Pathak, Elisa Nguyen, Michelle Peters, Yasmin Schmitt, Jörg Schlötterer, Maurice van Keulen, Christin Seifert ACM CSUR 2023 [paper][website] |
A Hybrid Text Classification and Language Generation Model for Automated Summarization of Dutch Breast Cancer Radiology Reports Elisa Nguyen*, Daphne Theodorakopoulos*, Shreyasi Pathak, Jeroen Geerdink, Onno Vijlbrief, Maurice van Keulen, Christin Seifert CogMI 2020 [paper] |
Pokérator - Unveil your inner Pokémon Dominique Geissler*, Elisa Nguyen*, Daphne Theodorakopoulos*, Lorenzo Gatti ICCC 2020 [paper][code] |
Activities
Academic
- Service
- Reviewing: NeurIPS'24, eXCV@ECCV'24
- Workshop organisation: Volunteer at WiML@ICML'24, Co-organising the ATTRIB workshop @NeurIPS'24
- Teaching and supervision
- Since Jul 2024: Master thesis supervision for Albert Catalan.
- Since Mar 2024: Master thesis supervision for Anastasiia Alekseeva.
- Nov 2023 - Jul 2024: Research project supervision for Albert Catalan.
- Jun 2023 - Mar 2024: Master thesis supervision for Evgenii Kortukov. Read about his work here.
- WS 2022/2023: Tutor for Trustworthy Machine Learning lecture at University of Tübingen.
- SS 2021: TA Data Science at University of Twente.
- WS 2020/2021: TA Natural Language Processing at University of Twente.
- Q3 2016: Tutor Discrete Mathematics 2 at Nordakademie.
- Q1 2015: Tutor Discrete Mathematics 1 at Nordakademie.
- Misc
- PhD representative in the Tübingen AI Center Council starting September 2024.
- Attended OxML 2023 summer school in MLxFundamentals and MLxHealth.
Volunteering
Matters that are close to my heart are education, equal opportunity, and the ongoing climate crisis. I am a member of KI macht Schule whose mission it is to bring AI education to local schools in Germany, and Viva con Agua, where we raise funds for WASH (water, sanitation and hygiene) projects in partner countries in the global south. It's a lot of fun and I encourage anyone to try out volunteering!