Valentin Lafargue
PhD student at IMT, INRIA, ANITI 2 and IRIT focused on AI evaluation and auditing.
I'm passionate about interdisciplinary work, exploring how AI intersects with fields like Law, Sociology, Cultural Analytics, Psychology, and Education.
Publications
Google Scholar profile-
Probing Cultural Signals in Large Language Models through Author Profiling
Conference paper Empirical Methods in Natural Language Processing (EMNLP) 2026
-
Exposing the Illusion of Fairness: Auditing Vulnerabilities to Distributional Manipulation Attacks
Conference paper ECML PKDD Applied Data Science Track 2026
-
Fairness is in the details: Face Dataset Auditing
Conference paper ECML PKDD Applied Data Science Track 2025
-
Orthogonium: A Unified, Efficient Library of Orthogonal and 1‑Lipschitz Building Blocks
Workshop paper Championing Open-source DEvelopment in ML Workshop @ ICML25
Education
-
2024 – present
PhD in Mathematics
Disloyalty and Biases in Generative AI
IMT, INRIA, ANITI 2, IRIT
- Supervised by J.-M. Loubes and E. Claeys
- Research on fairness, robustness and bias auditing in machine learning and generative AI
- Teaching duties: Lecture, project/TER supervision, tutorials (TDs) and practicals (TPs)
-
2023 – 2024
Master 2 (MSc)
Data Science and Data Engineering
Université Paul Sabatier, Toulouse
Deep Learning, NLP & SVACS, advanced statistics, parallel databases and cloud computing, business management, NoSQL, ethical communication, law for data.
-
2022 – 2023
Master 1
Data Science and Data Engineering
Université Paul Sabatier, Toulouse
Statistics, optimization, time series, machine learning, data warehousing, data mining, advanced algorithmics, advanced Python.
-
2019 – 2022
Bachelor
Mathematics, Higher Education and Research
Université Paul Sabatier, Toulouse
Teaching
-
2026–2027
- M1 SID UTMath methods for supervised Machine Learning TDs/TPs
- M2 SID UTDeep Learning Lectures/TDs/TPs
- M2 Mapi3 & IMA UTMath for Machine Learning TDs/TPs
-
2025–2026
- 1st year INSAMaths 0 TDs (Logic, Set theory, Applications, Sequence)
- M1 SID UTMath methods for supervised Machine Learning TDs/TPs
- M2 Mapi3 & IMA UTMath for Machine Learning TDs/TPs
-
2024–2025
- M1–M2 SID UT3Inter-promotional project supervision.
- M1 SID UT3TER supervisions, M1 students working on ML in education
- M2 SID UT3Learning process for Big Data, presentation jury