Characterizing out-of-distribution generalization of neural networks: application to the disordered Su–Schrieffer–Heeger model
Year: 2024
Type: article
Abstract: Machine learning (ML) is a promising tool for the detection of phases of matter. However, ML models are also known for their black-box construction, which hinders understanding of what they learn from the data and makes their application to novel data risky. Moreover, the central challenge of ML is to ensure its good generalization abilities, i.e. good performance on data outside the training set. Here, we show how the informed use of an interpretability method called class activation mapping, and the analysis of the latent representation of the data with the principal component analysis can increase trust in predictions of a neural network (NN) trained to classify quantum phases. In particular, we show that we can ensure better out-of-distribution (OOD) generalization in the complex classification problem by choosing such an NN that, in the simplified version of the problem, learns a known characteristic of the phase. We also discuss the characteristics of the data representation learned by a network that are predictors of its good OOD generalization. We show this on an example of the topological Su–Schrieffer–Heeger model with and without disorder, which turned out to be surprisingly challenging for NNs trained in a supervised way. This work is an example of how the systematic use of interpretability methods can improve the performance of NNs in scientific problems.
Language: English
Subfield: Materials Chemistry
Field: Materials Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,610
Funders: Simons Foundation, “la Caixa” Foundation, Centres de Recerca de Catalunya, Institut de Ciències Fotòniques, Flatiron Health, European Commission, Fundacja na rzecz Nauki Polskiej, Generalitat de Catalunya, Narodowe Centrum Nauki, Barcelona Supercomputing Center, Infrastruktura PL-Grid, QuantERA, European Climate, Infrastructure and Environment Executive Agency, HORIZON EUROPE Framework Programme, Agència de Gestió d'Ajuts Universitaris i de Recerca, European Regional Development Fund, Agencia Estatal de Investigación, Academic Computer Centre Cyfronet, AGH University of Science and Technology, H2020 European Research Council, HORIZON EUROPE European Innovation Council
Awards: G1049070721, G1188902550, G120905174, G1255325603, G1272927172, G148743912, NExt level Quantum information processing for Science and Technology, G1622782223, G162601807, G1676550097, Junior Leader la Caixa Postdoctoral Fellowship Programme: Shaping the new generation of leaders in research, G1908469581, Programmable Atomic Large-scale Quantum Simulation 2 - SGA1, G2105798022, G2236380367, G2262748287, G2575991346, G258593807, G2744780120, G2763573136, G2871827553, G2988934680, G3005324705, G3323404375, G3480869486, G348919436, G357246561, G3639969887, G3674674691, G3681454997, G3798495387, G3939077204, G3970178747, G4042783231, G408490004, G4239348134, G425668148, G434337789, G4359800586, G4368799831, G4391008399, G451917667, G4630312626, G4680432388, G4727238542, G4993383890, G5065677113, G5080475149, G531036588, G5426121769, G5587492786, G5615852374, G5620853212, G575169067, G5852091828, QuantERA II ERA-NET Cofund in Quantum Technologies, G5955501538, G5967599077, G598773271, G5987792135, G6145842628, G6166689270, G6275411036, G6364969246, G6482940084, G6552004136, G6656126374, G6727213032, G6731467812, G6743926034, G6835105602, G704744937, G7048927246, G7084143925, G7327758667, G7425137858, G7613962116, G7731791180, G784088340, G8023785571, G8180681537, G8283979283, Optical Topologic Logic, G8665153912, G8684520763, G8774671323, G8959349008
OpenAlex ID: W4404167686