Publication

Geometry of the Loss Landscape in Overparameterized Neural Networks: Symmetries and Invariances

Related publications (30)

Critical properties of the quantum Ashkin-Teller chain with chiral perturbations

We investigate the nature of the phase transitions in the quantum Ashkin-Teller chain in the presence of chiral perturbations. We locate the Lifshitz line separating a region of direct chiral transitions from the region where the transition is through an i ...

Amer Physical Soc2023

Retraction-based numerical methods for continuation, interpolation and time integration on manifolds

Axel Elie Joseph Séguin

The goal of this thesis is the development and the analysis of numerical methods for problems where the unknown is a curve on a smooth manifold. In particular, the thesis is structured around the three following problems: homotopy continuation, curve inter ...

EPFL2023

The kinectome: A comprehensive kinematic map of human motion in health and disease

Enrico Amico, Emahnuel Troisi Lopez, Antonella Romano

Human voluntary movement stems from the coordinated activations in space and time of many musculoskeletal segments. However, the current methodological approaches to study human movement are still limited to the evaluation of the synergies among a few body ...

WILEY2022

An automatized signal timing detection at signalized intersections with detailed vehicle trajectory data

Nikolaos Geroliminis, Jasso Espadaler Clapés, Emmanouil Barmpounakis

Signalized intersections are a fundamental part of urban networks. Their understanding is crucial to identify congestion patterns, queues, delays, and safety issues in local and network level. In this work, we analyze multimodal vehicle trajectories and pr ...

2022

Investigating cognitive ability using action-based models of structural brain networks

Enrico Amico

Recent developments in network neuroscience have highlighted the importance of developing techniques for analysing and modelling brain networks. A particularly powerful approach for studying complex neural systems is to formulate generative models that use ...

OXFORD UNIV PRESS2022

Information Retrieval Under Network Uncertainty: Robust Internet Ranking

Ralf Seifert, Anna Timonina-Farkas

Internet ranking algorithms play a crucial role in information technologies and numerical analysis due to their efficiency in high dimensions and wide range of possible applications, including scientometrics and systemic risk in finance (SinkRank, DebtRank ...

2023

Applicability of the landscape evolution model in the absence of rills

David Andrew Barry, Paolo Perona, Seifeddine Jomaa, Andrea Rinaldo, Mohsen Cheraghi, Andrea Cimatoribus

Despite numerous applications of physically-based models for incised landscapes, their applicability for overland flow on unchanneled surfaces is not known. This work challenges a widely used landscape evolution model for the case of non-uniform rainfall a ...

2022

Maintenance scheduling of manufacturing systems based on optimal price of the network

Olga Fink

Goods can exhibit positive externalities impacting decisions of customers in social networks. Suppliers can integrate these externalities in their pricing strategies to increase their revenue. Besides optimizing the prize, suppliers also have to consider t ...

2022

Supplementary datasets for "ARBRE: Computational resource to predict pathways towards industrially important aromatic compounds"

Vassily Hatzimanikatis, Homa Mohammadi Peyhani, Anastasia Sveshnikova

Supplementary datasets accompanying the manuscript "ARBRE: Computational resource to predict pathways towards industrially important aromatic compounds" published in the Metabolic Engineering Journal (https://doi.org/10.1016/j.ymben.2022.03.013). In line w ...

Zenodo2022

Finite Littlestone Dimension Implies Finite Information Complexity

Michael Christoph Gastpar, Aditya Pradeep, Ido Nachum

We prove that every online learnable class of functions of Littlestone dimension d admits a learning algorithm with finite information complexity. Towards this end, we use the notion of a globally stable algorithm. Generally, the information complexity of ...

2022