AI Research Intern & Applied Mathematics Student.
CMAP at Ecole Polytechnique, France
Focusing on AI Safety, Robustness, and Statistical Learning.
About Me —
Engineering & Labs —
Uses —
Hobbies
LinkedIn —
Twitter —
Resume
-
ArtDanjProduction
- Remote
-
06:49
(UTC +02:00) - https://go.arthurdanjou.fr/website
- https://go.arthurdanjou.fr/twitter
- https://go.arthurdanjou.fr/linkedin
- https://go.arthurdanjou.fr/discord
Highlights
Pinned Loading
-
-
bibliography
bibliography PublicCategorized BibTeX bibliography on Byzantine-robust distributed and federated learning, for use with biblatex and biber.
TeX 1
-
DropoutReducesUnderfitting
DropoutReducesUnderfitting PublicTensorFlow/Keras implementation of "Dropout Reduces Underfitting" (Liu et al., 2023). A comparative study of Early and Late Dropout strategies to optimize model convergence.
Jupyter Notebook
-
Learning-Theory-Two-layer-Neural-Networks
Learning-Theory-Two-layer-Neural-Networks PublicGeneralization bounds for two-layer ReLU neural networks via scale-invariant complexity measures
TeX
-
ai-sync-dotfiles
ai-sync-dotfiles PublicOne source of truth for MCP servers, skills and slash commands across Claude Code, OpenCode, Codex, Zed and more — plus classic dotfiles.
TypeScript
If the problem persists, check the GitHub status page or contact support.





