Various projects I worked on
Overview of multiple projects, some of them have their dedicated page you can access to see demos (under construction).
CharlesGPT — home-made documented chatbot
End-to-end project to explore how a small generative model works in practice: Flutter client, FastAPI backend, local GPT-2 inference, token-level metrics (top‑k, perplexity) and adjustable sampling parameters.
CharlesGPT demo.
Identifying codes in graphs: study and coding experiments
2-month research project on identifying codes in graphs, based on the paper Charon et al., Minimum Sizes of Identifying Codes in Graphs Differing by One Vertex.
Coding experiments and implementations: Identifying_codes_graphs.
Poster: view.
Experiments for propagation of chaos in mean field underdamped regime for wide neural networks
5-month internship at RIKEN AIP (Deep Learning Theory Team).
Developed MFLD-PoC notebooks and related simulation tooling for mean-field proof-of-concept.
You can find the experiments here: MFLD-PoC.
A report on the project (2nd part): view.
ModularyNN — modular neural network framework (Python)
Lightweight, composable neural network building blocks for experiments and teaching: customizable layers and training loops from scratch.
Repository: modularyNN.
Technical support for Ardian startup studio
3-day Ardian Startup Studio hackathon.
Technical support for prototyping ideas for internal use at Ardian France; demonstrated feasibility and identified long-term limitations.
Yearly event in Paris (see Artefact press release).
KIRO Inter School Operational Research Challenge
3 participations with strong results on optimization problems from RTE, Renault, and Pelico.
See KIRO challenge.
Repository from first participation (extended): Projet_RO_RTE.
AI and machine learning for theme park attraction attendance prediction
One-day challenge, ranked 2nd at Ecole des Ponts. Department project with Eleven Strategy consulting.
See our slides: view.
Simulation and learning for gas storage refueling optimization
6-month project at Ecole des Ponts in partnership with Air Liquide R&D. Superior cost function results compared to previous naive methods.
Poster: view.
You can see our report here: view.
Differential and Riemannian geometry — general relativity and black hole trajectory simulation
6-month research project (1st year, Ecole des Ponts).
Reproducing results from generic simulations. Supervised by Éric Cancès.
You can see our report here: view.
Modeling granular media using cellular automata
One and a half year project as part of mathematical and physics preparatory classes (MP).
Mathematical theory of cellular automata and working with open-source hashing code to reproduce erosion and craterization with custom rules (using Golly).
See description.