"Where there is matter data, there is geometry"

Kepler

Giovanni Luca Marchetti

I am a postdoctoral researcher at the Royal Institute of Technology (KTH) in Stockholm, Sweden.


My research focuses on geometric approaches to machine learning and high-dimensional statistics. More specifically, I am interested in algebro-geometric aspects of deep neural networks, manifold/representation learning, geometric density estimation, and topological data analysis.

Email     X/Twitter     GitHub     LinkedIn

Research

Below you can find a selection of my academic works, subdivided into topics. Click on an image to access the arXiv entry. For a complete list, please visit my Google Scholar profile.


Equivariant/Invariant Deep Learning: These works explore the interaction between symmetry and deep learning, ranging from the invariant theory of neural networks to equivariant representation learning, with applications to robotics.

Harmonics of Learning: Universal Fourier Features Emerge in Invariant Networks

COLT 2024

Equivariant Representation Learning via Class-Pose Decomposition

AISTATS 2023

Equivariant Representation Learning in the Presence of Stabilizers

ECML-PKDD 2023

Learning Geometric Representations of Objects via Interaction

ECML-PKDD 2023

Back to the Manifold: Recovering from Out-of-Distribution States

IROS 2022

Neural Lattice Reduction: A Self-Supervised Geometric Deep Learning Approach

TMLR 2025


Computational Geometry: These works concern high-dimensional Voronoi tessellations and Delaunay triangulations, with applications to density estimation and active learning.

An Efficient and Continuous Voronoi Density Estimator
AISTATS 2023

Voronoi Density Estimator for High-Dimensional Data: Computation, Compactification and Convergence

UAI 2022

Active Nearest Neighbor Regression Through Delaunay Refinement

ICML 2022

Hyperbolic Delaunay Geometric Alignment

ECML-PKDD 2024

HyperSteiner: Computing Heuristic Hyperbolic Steiner Minimal Trees

ALENEX 2025

Algebraic Geometry of Deep Learning: These works explore the (algebraic) geometry of function spaces defined by neural networks.

Geometry of Lightning Self-Attention: Identifiability and Dimension

ICLR 2025

On the Geometry and Optimization of Polynomial Convolutional Networks

AISTATS 2025

An Invitation to Neuroalgebraic Geometry

Preprint


Other: These works concern various topics in pure mathematics, e.g., category theory and combinatorics.

Hearts and Towers in Stable -Categories

Journal of Homotopy 14

Rubik's Abstract Polytopes

Preprint


Thesis: I obtained my doctoral degree from KTH in 2024 under the supervision of Prof. Danica Kragic. Below you can download the thesis and the accompanying slides.

On Symmetries and Metrics in Geometric Inference

KTH 2024
Slides

Resume


Click here to download my academic resume.