Skip to content
View Dorianhgn's full-sized avatar

Block or report Dorianhgn

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Dorianhgn/README.md

Hi there 👋 I'm Dorian Hugonnet

🎓 I'm a final-year Engineering Student in Applied Mathematics at INSA Toulouse, specializing in advanced Artificial Intelligence architectures and their practical, scalable applications.

🚀 My core passion lies in designing and operationalizing cutting-edge deep learning solutions that are both high-performing and interpretable, bridging theoretical advances with robust production systems.

My Key Areas of Focus & Expertise:

  • 🧠 High-Performance Deep Learning Architectures: Deeply engaged with state-of-the-art models like Mamba State Space Models (SSMs) for efficient sequence modeling, and exploring innovative generative techniques such as Conditional Flow Matching (CFM) for complex data generation and analysis.
  • ⚙️ Scalable Model Development & Deployment: Focused on building production-ready AI systems, leveraging advanced Hyperparameter Optimization (HPO) strategies (e.g., using Optuna) and developing robust REST APIs with FastAPI for seamless, efficient integration.
  • 💡 Explainable AI (XAI): Committed to ensuring transparency and understanding in complex AI models, making them trustworthy and actionable.

Research & Applied Projects:

  • 🔬 Atmospheric Dust Plume Prediction (Mamba & CFM): Developed novel, efficient deep learning models using Mamba SSMs and Conditional Flow Matching for 72-hour dust plume evolution prediction. This cutting-edge research project is currently being prepared for academic publication, hence the repository is not publicly available yet.
  • 📚 Multi-Level Variance Estimation in Log-Euclidean Geometry (CERFACS): Collaborated with CERFACS Toulouse on a research project, culminating in a scientific publication draft. This experience solidified my interest in rigorous scientific inquiry and advanced mathematical methodologies. Paper here
  • 🧠 Deep Dive into Transformer Architectures: Rebuilt a Transformer from scratch to gain a fundamental understanding of LLM internals and their operational principles. Repository
  • 🔍 Interpretability in Machine Learning: Applied concepts from Interpretable Machine Learning in a course project, exploring various methods for enhancing model transparency and understanding. Repository

Beyond the Code:

🏃‍♂️ I'm also a national-level high jump athlete, a pursuit that has instilled in me discipline, resilience, and a relentless drive for continuous performance improvement. I'm always curious about the intersection of AI and sports.


📬 Feel free to connect with me on LinkedIn!

Pinned Loading

  1. Log-Euclidean-MLMC-Estimator Log-Euclidean-MLMC-Estimator Public

    Dans le cadre de notre projet d'initiation à la recherche en 4e année à l'INSA de Toulouse, nous avons étudié l'estimateur de la covariance en géométrie log-euclidienne.

    Jupyter Notebook

  2. ProjetML4A ProjetML4A Public

    Jupyter Notebook 1