Publication

Benchmarking machine-readable vectors of chemical reactions on computed activation barriers

Publications associées (31)

NMR Crystallography in the Big Data Era: New Methods and Applications Powered by Machine Learning

Structure determination of materials is key to understanding their physical properties. While single-crystal X-ray diffraction is the gold standard for structures displaying long-range order, many materials of interest are polycrystalline and/or disordered ...

EPFL2023

Nonlinear dynamics and Kerr frequency comb formation in lattices of coupled microresonators

Tobias Kippenberg, Kenichi Komagata, Alexey Tikan, Aleksandr Tusnin

Recently, substantial progress has been made in the understanding of microresonators frequency combs based on dissipative Kerr solitons (DKSs). However, most of the studies have focused on the single-resonator level. Coupled resonator systems can open new ...

Nature Portfolio2023

Snap buckling of bistable beams under combined mechanical and magnetic loading

Pedro Miguel Nunes Pereira de Almeida Reis, Dong Yan, Tomohiko Sano, Arefeh Abbasi

We investigate the mechanics of bistable, hard-magnetic, elastic beams, combining experiments, finite-element modelling (FEM) and a reduced-order theory. The beam is made of a hard magneto-rheological elastomer, comprising two segments with antiparallel ma ...

ROYAL SOC2023

De novo designed proteins: a study in engineering novel folds and functions

Alexandra Krina Van Hall-Beauvais

Proteins control nearly every facet of life on a molecular level. Proteins are formed from linear strings of amino acids, which fold into three-dimensional structures that can enact functions. Evolution has created highly efficient proteins in diverse fold ...

EPFL2023

On the symmetries in the dynamics of wide two-layer neural networks

Lénaïc Chizat

We consider the idealized setting of gradient flow on the population risk for infinitely wide two-layer ReLU neural networks (without bias), and study the effect of symmetries on the learned parameters and predictors. We first describe a general class of s ...

AMER INST MATHEMATICAL SCIENCES-AIMS2023

Hypervalent Iodine: New Reagents and Functionalization of Peptides

Eliott Hugo Joran Le Du

Alkynes are found in a multitude of natural or synthetic bioactive compounds. In addition to the capacity of these chemical motifs to impact the physicochemical properties of a molecule of interest, the well-established reactivity of alkynes makes them ide ...

EPFL2023

Tungsten oxide-based photoanodes: biomass valorization and the effects of oxygen vacancies

Nukorn Plainpan

Tungsten oxide-based materials have unique and intriguing properties as photoanode materials for solar energy conversion. The first part of the thesis focuses on the use of WO3 photoanodes for the oxidation of 5-(hydroxymethyl)furfural (HMF). Chapter one l ...

EPFL2023

A Geometric Transformer for Structural Biology: Development and Applications of the Protein Structure Transformer

Lucien Fabrice Krapp

Proteins, the central building blocks of life, play pivotal roles in nearly every biological function. To do so, these macromolecular structures interact with their surrounding environment in complex ways, leading to diverse functional behaviors. The predi ...

EPFL2023

Long-range electrostatic contribution to electron-phonon couplings and mobilities of two-dimensional and bulk materials

Massimiliano Stengel, Nicola Marzari, Marco Gibertini, Samuel Poncé

Charge transport plays a crucial role in manifold potential applications of two-dimensional materials, in-cluding field-effect transistors, solar cells, and transparent conductors. At most operating temperatures, charge transport is hindered by scattering ...

AMER PHYSICAL SOC2023

Reduced representations of complexes, signals, and multifiltrations

Celia Camille Hacker

The field of computational topology has developed many powerful tools to describe the shape of data, offering an alternative point of view from classical statistics. This results in a variety of complex structures that are not always directly amenable for ...

EPFL2022

SPECTRE: Spectral Conditioning Helps to Overcome the Expressivity Limits of One-shot Graph Generators

Nathanaël Perraudin, Andreas Loukas, Karolis Martinkus

We approach the graph generation problem from a spectral perspective by first generating the dominant parts of the graph Laplacian spectrum and then building a graph matching these eigenvalues and eigenvectors. Spectral conditioning allows for direct model ...

JMLR-JOURNAL MACHINE LEARNING RESEARCH2022