CV

===

Education

• Toulouse Mathematical Institute (non-curricular), Sep 2024 – June 2025
  • Open-access seminars for researchers
  • Deep learning, probability, optimization, statistics
• M.S. in Applied Mathematics & Engineering, National Institute of Applied Sciences (INSA Toulouse), Sep 2021 – Sep 2025
  • Relevant coursework: High-Dimensional Deep Learning, AI Frameworks, Signal & Image Processing, Machine Learning, Data Analysis, Statistical Modeling, Continuous Optimization, Non-Differentiable Optimization, Advanced Probability, Stochastic Processes
• Classe préparatoire aux grandes écoles, Fabert High School, Sep 2018 – Jun 2021
  • Major in Physics & Chemistry
  • Minor in Mathematics & Informatics
  • Honorary class

Work experiences

• June 2026 – Now: Research Scientist
  • Deezer Research
  • Paris, France
• March 2025 – January 2026: Research Intern
  • ADASP (Télécom Paris) x Deezer Research
  • Memorization detection and application to the detection of copyrighted data in audio generation
  • Duties included:
    • Critical analysis and improvement of the MTG laboratory’s MIRA tools for data replication detection
    • Development of Membership Inference Attacks on music Flow Matching models exploiting memorization
  • Supervisors: G. Peeters, G. Meseguer-Brocal
• July – December 2023: Neural Network Internship
  • ONERA, the French Aerospace Lab
  • Toulouse, France
  • Duties included:
    • Theoretical conceptualization and computational development of a mixed-weight neuro-fuzzy network
    • Development of a binary & continuous optimizer
    • Development from scratch without using deep learning/machine learning libraries
  • Supervisor: J.-L. Farges

Volunteer experiences

• February – July 2024: Civic Service in Agroecology
  • Djarama, Senegal
  • Duties included: Vegetable cultivation, raising environmental and nutritional awareness among children
• September 2022 – 2024: Member of the Strategic Orientation Committee
  • Junior INSA Service (School’s junior company)
  • Duties included: Guidance on Junior Enterprise strategy, internal audit of treasury department
• September 2022 – 2024: Treasurer
  • Junior INSA Service (School’s junior company)
  • Duties included: Accounting and financial management, creation and delivery of treasury training courses

Side Projects

• Pyzyconfig [In progress] — Python Framework
  • Easy-to-use configuration saver for experiments in Python. Useful for small projects that do not require a complex Hydra pipeline.
• EasyStart
  • Small pipelines that automate commonly used work configurations with a single keystroke (VPN detection required, SSH mounting, opening specific applications, etc.)

Skills & certification

• Programming Languages & tools
  • Python (PyTorch, TensorFlow, scikit-learn)
  • bash, git
• Technical content writing
• Languages:
  • French (Native)
  • English (TOEIC score 975/990)

Publications

Where Flow Matching Leaks: Characterising the Membership Signal Along the Interpolation Path Permalink

sesmat2026where, title={Where Flow Matching Leaks: Characterising Membership Signals Along the Interpolation Path}, author={Thomas Sesmat and Gabriel Meseguer-Brocal and Geoffroy Peeters}, booktitle={Forty-third International Conference on Machine Learning}, year={2026}, url={https://openreview.net/forum?id=Ty5X41WbJw}

Neuro-fuzzy networks for inference transparency Permalink

Sesmat, T., Farges, J.-L., Perotto, F., and Picard, G (2024). " Réseaux neuro-flous pour la transparence de l’inférence "Journées d'Intelligence Artificielle Hybride : de l'intégration des connaissances et de l'humain à l'explication des modèles (HyCHA'24)."

Award

• 1st place, Data Challenge IA Pau, 2023 (€3000 cash prize)
  • Conceptualization and development of a technical solution for a speech and video analysis tool