The DOI matches its Crossref record; 112 of the 112 references it cites were found in a registry or an index; none of the 68 sources its claims cite is checked for support yet.
Unsupervised machine learning in atomistic simulations, between predictions and understanding
Open accessAbstract
Automated analyses of the outcome of a simulation have been an important part of atomistic modeling since the early days, addressing the need of linking the behavior of individual atoms and the collective properties that are usually the final quantity of interest. Methods such as clustering and dimensionality reduction have been used to provide a simplified, coarse-grained representation of the structure and dynamics of complex systems from proteins to nanoparticles. In recent years, the rise of machine learning has led to an even more widespread use of these algorithms in atomistic modeling and to consider different classification and inference techniques as part of a coherent toolbox of data-driven approaches. This perspective briefly reviews some of the unsupervised machine-learning methods—that are geared toward classification and coarse-graining of molecular simulations—seen in relation to the fundamental mathematical concepts that underlie all machine-learning techniques. It discusses the importance of using concise yet complete representations of atomic structures as the starting point of the analyses and highlights the risk of introducing preconceived biases when using machine learning to rationalize and understand structure-property relations. Supervised machine-learning techniques that explicitly attempt to predict the properties of a material given its structure are less susceptible to such biases. Current developments in the field suggest that using these two classes of approaches side-by-side and in a fully integrated mode, while keeping in mind the relations between the data analysis framework and the fundamental physical principles, will be key to realizing the full potential of machine learning to help understand the behavior of complex molecules and materials.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
This perspective reviews unsupervised machine-learning methods for atomistic simulations—clustering and dimensionality reduction—and argues they are powerful but strongly dependent on structural representations and hyperparameter choices. It recommends validating and optimizing unsupervised analyses through supervised learning tasks to reduce preconceived biases and improve understanding of structure–property relations.
Key findings
- Unsupervised schemes such as clustering and dimensionality reduction can identify structural motifs and coarse-grained states, but their outcome depends crucially on the feature space, metric, or kernel; non-linear transformations can split, merge, create, or destroy local maxima in the probability distribution.
- The paper states that for real data there is no “ground truth” for what partitioning or low-dimensional representation is best, and analyses often require human intervention in algorithm design and hyperparameter tuning, risking tuning until results match expectations.
- Supervised learning can provide objective metrics for unsupervised representations: checking how accurately a property can be predicted, analyzing learning curves, optimizing hyperparameters, and benchmarking order parameters against the committor in biomolecular simulations.
- The paper argues that combining supervised and unsupervised approaches side-by-side is key, with examples including Markov-state models, materials cartography, accelerated sampling, and structure searches; no original numerical results are reported.
Methods
- Perspective/review article; no new primary data, simulations, or experiments are reported.
- Reviews mathematical foundations of representations, distances/kernels, clustering, dimensionality reduction, and applications in atomistic modeling.
- Synthesizes literature and author opinion rather than performing a systematic quantitative meta-analysis.
Limitations
- The paper states that unsupervised learning lacks ground truth for real data and is sensitive to representation and hyperparameters; manual intervention can introduce preconceived biases.
- The paper states that supervised validation is proposed as a mitigation, but this is presented as the author’s opinion and not as a fully solved problem.
- Evident from the text: as a perspective, it provides no new numerical benchmarks or empirical validation, and its conclusions are based on literature synthesis and the author’s viewpoint.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 of 68 cited sources checked · 0 supported · 0 flagged
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In all honesty, most of the time the analyses did require a significant amount of human intervention, first in the design of the details of the classification/dimensionality reduction algorithm, and later in the choice of the hyperparameters that define the structural representation and tune the functioning of the machine learning scheme.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes? Based: unclearIn the paper
“When considering the nitrogen-substituted analogues of pentacene, however, we found that the most clear-cut clustering was associated with in-plane H-bonding rather than with the stacking.”
What the sources say
- Not checked yetcites [84] 2018, Chem. Sci. source
Strength:boosted· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIn the paper
“we have recently developed a generalized convex hull (GCH) construction 113 that uses (kernel) principal component analysis to determine the structural parameters that capture the highest degree of diversity in a collection of (meta)stable structures.”
What the sources say
- Not checked yetcites [113] A. Wieczorek , M.Wieser, D.Murezzan, and V.Roth, “Learning sparse latent representations with the deep copula information bottleneck,” e-print arXiv:1804.06216 (2018).
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesA conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Supervised machine-learning techniques, that explicitly attempt to predict the properties of a material given its structure, are less susceptible to such biases.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesThe paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“A second reason - that is particularly relevant for unsupervised-learning applications geared towards simplifying the description of complex systems - is that the choice of representation provides ample leeway to incorporate prior knowledge or to otherwise restrict the scope of the analysis.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesThe paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The main downside with using raw atomic coordinates as an input representation is precisely the fact that they discriminate between structures that are physically indistinguishable.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In closing this brief overview, it should be stressed that all of these techniques depend crucially on the choice of the feature space, and/or the metric that is used to estimate the similarity between inputs.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Machinelearning representations that are based on the description of local motifs, rather than on a conventional characterization of the system in terms of its lattice parameters and minimal unit cell, are very convenient, because they simplify the task of recognizing structures that correspond to alternative descriptions of the same periodic configuration (e.g. a conventional versus a primitive f”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIn the paper
“While the cost and effectiveness of different representations may vary, one can demonstrate that most of these schemes can be seen as a specific implementation of a very abstract density-based scheme, in which a ket |A〉 is associated to each configuration 37 ,”
What the sources say
- Not checked yetcites [37] 2019, J. Chem. Phys. source
Strength:boosted· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesA conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The connection between density peaks and minima in the free energy makes density-based clustering methods (Fig. 3b) particularly well-suited to the analysis of finitetemperature atomistic simulations.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes
- It may state two separate claims as one. 14
- Probably not a substantial claim. 30
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In other terms Eq. (1) reproduces exactly the value of the kernel for each structure in the reference set, and provides a practical approximation of the vector |X〉 associated with the kernel for any other structure.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes? Based: unclearIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Obviously, opting for a specialized description of a system introduces some biases in the procedure.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“From a formal point of view, the representer theorem guarantees that each positive-definite kernel 14 is associated with a Hilbert space whose scalar product 〈X|X ′ 〉 equals the kernel, and it is always possible to convert a kernel into a metric, e.g. by the relation d ( X , X ′ ) 2 = k ( X , X ) + k ( X ′ , X ′ ) -2 k ( X , X ′ ).”
What the sources say
- Not checked yetcites [14] M. Cuturi , e-print arXiv:0911.5367 (2009).
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“It is worth stressing that the representation of a structure can take the form of a vector of features |X〉 (e.g. all the backbone dihedrals in a protein, the coordination number of atoms in a cluster, etc.), a measure of similarity between pairs of structures that can be implemented as a distance d ( X , X ′ ), or the form of a kernel k ( X , X ′ ). In fact, these three ways of representing a data”
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Given a representation in terms of a finite set of features x j ( X ) = 〈 j |X〉 , it is always possible to use it as the starting point to define a distance or a kernel.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Such a unifying view of input representations helps to understand the limits and strengths of each particular representation - for instance how SOAP power spectra components are systematic but overdetermined, and how Behler-Parrinello symmetry functions are highly correlated with each other.”
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“It is however hard if not impossible to comprehend - not to mention visualize - the meaning or the relative positions of vectors in a high (or infinite) dimensional space.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✗ Checkable: fails✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“The problem of obtaining a complete yet concise description of a molecular or condensed-phase structure is central to many modelling techniques, including the sampling of rare events 15 and the search for stable structures 2,16,17 .”
What the sources say
Strength:plain· not yet calibratedTests✓ Asserted: passes✗ Checkable: fails✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“One may argue that given sufficient amounts of training data a good ML scheme should be able to recognize these invariances, but why bother: most of the state-ofthe-art representations incorporate symmetries and by using them you can let your 20-layer convolutional neural network focus on learning correlations that are not as obvious.”
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Clustering techniques identify agglomeration of data in the most significant regions of feature space, but do not offer an overall picture of the relations between different structures.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In the context of unsupervised learning it is often performed manually, by trial and error, and constitutes one of the critical points to consider to keep the procedure truly unbiased.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“This is a problem that obviously transcends the specific application to chemistry, physics or materials science, and dimensionality-reduction methods have long been sought after for all sorts of data analysis applications.”
Strength:plain· not yet calibratedTests✓ Asserted: passes✗ Checkable: fails✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“Obviously, the separation is far from clear-cut, and it is often possible to construct a local version of a global representation 25 , and vice versa to assemble local kernels and feature vectors into a representation of entire structures 32 .”
What the sources say
Strength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The use of local, atom/fragment-based representations is also convenient to find recurring patterns in the building blocks of a structure, rather than in the overall configuration.”
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As we will discuss below, the risk of introducing spurious features can be mitigated by using general-purpose, relatively simple functions of the atomic coordinates as the input representation, and by checking that clustering is insensitive to the choice feature space, kernel or metric.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In principle, the information on the absolute position of each input in space, and the relations between any given pair of inputs, are encoded fully in the feature vectors.”
Strength:hedged· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Non-linear transformations of the input space, or of the measure of similarity, introduce distortions that can easily split, merge, create and destroy local maxima in the probability distribution associated with the input set.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“dimensionality reduction lies at the heart of the vast majority of accelerated sampling methods, which try to beat the time scale problem of molecular simulations”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
In the paper
“The idea that the data may lie on a manifold that is only low-dimensional on a local basis is also behind methods such as locally-linear embedding (LLE) 76”
What the sources say
- Not checked yetcites [76] 2000, Science source
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
None of the sources cited for this claim have been checked yet.
In the paper
“The time-lagged independent component analysis method (TICA) 80,81 identifies linear combinations of features from a time-dependent trajectory that not only have the largest variance, but that also have the longest autocorrelation time”
What the sources say
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Rests on an assumption: The most relevant degrees of freedom for a system are those in which the dynamics are slower.
In the paper
“the spectral gap optimization of order parameters (SGOOP) method to single out the linear combination of input features that correspond to the direction that is most relevant for the long-time dynamics of the system being studied 82”
What the sources say
- Not checked yetcites [82] 2016, Proc. Natl. Acad. Sci. U. S. A. source
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Rests on an assumption: The most relevant degrees of freedom for a system are those in which the dynamics are slower.
None of the sources cited for this claim have been checked yet.
In the paper
“'Materials cartography', as it has been referred to 57 , strives to obtain a low-dimensional (typically 2D) representation of a set of materials, or of different phases of the same material, that reflects the 'structureenergy-property' landscape 84,85 of that system.”
What the sources say
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“Similar approaches have been applied to biomolecular simulations 86 , where the main application of dimensionality reduction has been in connection with accelerated sampling.”
What the sources say
- Not checked yetcites [86] 2012, J. Mol. Graphics Modell. source
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“All of the methods discussed this far rely exclusively on the distribution of atoms in each structure.”
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“ISOMAP 75 uses a neighborhood analysis to estimate geodesic distances between points, based on the assumption that the data lies on a curved, but locally low-dimensional, manifold.”
What the sources say
- Not checked yetcites [75] 2000, Science source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The applications of dimensionality reduction schemes to atomistic simulations reflect the dual function of a coarse-grained representation, as a mean to simplify the description of a system to understand it better, and to explore more effectively its conformational space.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
None of the sources cited for this claim have been checked yet.
In the paper
“In order to observe a rare (activated) event without developing dedicated hardware 1 one needs to selectively increase the speed at which the pathways related to the transition are traversed 51,87,88 , to modify the probability distribution that is sampled during the simulation so that configurations that are close to the transition state occur more often 15,89 , or to generate adaptively ens”
What the sources say
- Not checked yetcites [1] 2009
- Not checked yetcites [51] 2002, Annu. Rev. Mater. Res. source
- Not checked yetcites [87] 2002, Annu. Rev. Phys. Chem. source
- Not checked yetcites [88] 2008, J. Phys. Chem. B source
- Not checked yetcites [15] 2002, Proc. Natl. Acad. Sci. U. S. A. source
- Not checked yetcites [89] 1977, J. Comput. Phys. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“Take for instance the case of metadynamics, in which a repulsive bias is accumulated in locations that have already been visited during the trajectory, forcing the system out of free energy minima 93,94 . The volume of one of these minima grows exponentially with the dimensionality d of collective variable space, and so does the time to accumulate sufficient bias to escape the freeenergy basin.”
What the sources say
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“Either way, the objective is to accelerate sampling while introducing a minimal disturbance to the natural behavior of the system 91”
What the sources say
- Not checked yetcites [91] 2013, Phys. Rev. Lett. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
None of the sources cited for this claim have been checked yet.
In the paper
“The accuracy of the regression model also provides an objective metric to optimize the hyperparameters that define the representation 107 , and even to determine the most effective non-linear function of features that reproduces the target properties 108 .”
What the sources say
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It is then tempting to pick one own's arbitrary perception of what an 'insightful' outcome of the analysis would be, and tune the representation and machine-learning scheme until the results reflect most clearly one's expectations.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✗ Checkable: fails✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may be too vague to check against evidence.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For starters, it gives an idea of the limiting accuracy that can be expected when inferring structureproperty relations by application of unsupervised classification or coarse-graining schemes.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
None of the sources cited for this claim have been checked yet.
In the paper
“Clustering can be used to select the most relevant reference configurations from a database 109 , or to partition it into regions on which specialized models can be trained. 110”
What the sources say
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“A low-dimensional representation of a dataset can help when it comes to identifying problematic regions, and understanding how a given representation 'sees' a set of structures 58,104,111 .”
What the sources say
Strength:plain· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Choosing to describe the inputs using a feature vector that is not problem specific, and validating and optimizing it through a relevant supervised learning exercise, reduces substantially the amount of human intervention and the possibility of introducing preconceived biases when applying clustering and dimensionality reduction schemes.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Unsupervised schemes can help shed light on the functioning of supervised algorithms that are often opaque, and supervised schemes can help to formulate quantitative metrics to assess the performance of unsupervised learning.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✗ Checkable: fails✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
The paper relies on this without support.
In biomolecular simulations, entropically-stabilized states are the norm rather than the exception.
“particularly in the case of biomolecular simulations in which entropically-stabilized states are the norm rather than the exception.”
The paper relies on this without support.
In crystalline materials with few defects, local minima on the potential energy surface unequivocally identify the meta-stable states of the system.
“This is because in crystalline materials with few defects local minima on the potential energy surface unequivocally identify the meta-stable states of the system.”
The paper relies on this without support.
The most relevant degrees of freedom for a system are those in which the dynamics are slower.
“It is often the case, however, that the most relevant degrees of freedom for a system are those in which the dynamics are slower.”
The paper relies on this without support.
The complexity of simulations is increasing for materials, chemical and biomolecular applications.
“as the complexity of simulations is increasing for materials, chemical and biomolecular applications.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“ML models, analysis of their key features (32,52), and dimensionality reduction (i.e., unsupervised learning) (53-56) represent important tools for building structure-property relationships and mapping chemical space (57).”
From Machine Learning for the Discovery, Design, and Engineering of Materials. (2022) · cites this paper as [56]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
PCA input vectors are described as d-dimensional
MinorThe text defines x_i as a D-dimensional input vector and s_i as its d-dimensional principal-component representation. It later calls the x_i vectors d-dimensional.
“Another way to interpret PCA is to see { x i } as the set of d -dimensional points”
“Given a set of D -dimensional input vectors x i associated with N structures or local environments, PCA identifies the d linear combinations of the feature components that capture the largest possible fraction of the variability of the dataset (Fig. 4a).”
LLE discussion points to the wrong figure panel
MinorThe LLE discussion points to Fig. 4c, but the caption assigns local-relation-preserving embedding methods to Fig. 4d; Fig. 4c is assigned to multidimensional scaling.
“and then by imposing self-consistently that each point in low dimension should be a combination of its neighbors using the same weights (Fig. 4c)”
“(c) Multi-dimensional scaling and related approaches attempt to reproduce the similarity between high-dimensional data points in low dimension. (d) Embedding methods explicitly try to preserve local relations between points, under the assumption that they lie on a (locally) low-dimensional manifold.”
Citations
Cites
(2008) · Journal of Machine Learning Research · Cited by 33,051
Nonlinear Dimensionality Reduction by Locally Linear Embedding
Roweis, S. T., & Saul, L. K. (2000) · Science · doi:10.1126/science.290.5500.2323 · Cited by 14,965
A Global Geometric Framework for Nonlinear Dimensionality Reduction
Tenenbaum, J. B., Silva, V. D., & Langford, J. C. (2000) · Science · doi:10.1126/science.290.5500.2319 · Cited by 13,585
Commentary: The Materials Project: A materials genome approach to accelerating materials innovation
Jain, A., Ong, S. P., Hautier, G., Chen, W., Richards, W. D., Dacek, S., Cholia, S., Gunter, D., Skinner, D., Ceder, G., et al. (2013) · APL Materials · doi:10.1063/1.4812323 · Cited by 13,094
Nonlinear Component Analysis as a Kernel Eigenvalue Problem
Schölkopf, B., Smola, A. J., & Müller, K. (1998) · Neural Computation · doi:10.1162/089976698300017467 · Cited by 8,196
Nonphysical sampling distributions in Monte Carlo free-energy estimation: Umbrella sampling
Torrie, G. M., & Valleau, J. P. (1977) · Journal of Computational Physics · doi:10.1016/0021-9991(77)90121-8 · Cited by 6,091
Laio, A., & Parrinello, M. (2002) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.202427399 · Cited by 5,626
Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces
Behler, J., & Parrinello, M. (2007) · Physical Review Letters · doi:10.1103/physrevlett.98.146401 · Cited by 5,010
Machine learning for molecular and materials science
Butler, K. T., Davies, D. W., Cartwright, H., Isayev, O., & Walsh, A. (2018) · Nature · doi:10.1038/s41586-018-0337-2 · Cited by 4,968
Clustering by fast search and find of density peaks
Rodríguez, Á., & Laio, A. (2014) · Science · doi:10.1126/science.1242072 · Cited by 4,963
Flagged first, then by citations
Research works printed but not linked: 7
- 2016, The God Delusion
- 2011, Pattern Recognition and Machine Learning (Information Science and Statistics)
- 2006, Gaussian Processes for Machine Learning
- 2003, Energy Landscapes: Applications to Clusters, Biomolecules and Glasses
- H. Huo and M.Rupp, e-print arxiv:1704.06439 (2017). · arXiv
- 2010, Multidimensional Scaling
- 2008, J. Mach. Learn. Res., 9, 2579
Cited by
Sorted by citations
Machine Learning Interatomic Potentials as Emerging Tools for Materials Science
Deringer, V. L., Caro, M. A., & Cśanyi, G. (2019) · Advanced Materials · doi:10.1002/adma.201902765 · Cited by 1,136
Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems
Keith, J. A., Vassilev-Galindo, V., Cheng, B., Chmiela, S., Gastegger, M., Müller, K.-R., & Tkatchenko, A. (2021) · Chemical Reviews · doi:10.1021/acs.chemrev.1c00107 · Cited by 849
Big-Data Science in Porous Materials: Materials Genomics and Machine Learning
Jablonka, K. M., Ongari, D., Moosavi, S. M., & Smit, B. (2020) · Chemical Reviews · doi:10.1021/acs.chemrev.0c00004 · Cited by 656
Physics-Inspired Structural Representations for Molecules and Materials
Musil, F., Grisafi, A., Bartók, A. P., Ortner, C., Cśanyi, G., & Ceriotti, M. (2021) · Chemical Reviews · doi:10.1021/acs.chemrev.1c00021 · Cited by 594
Self-citationPhysics-Inspired Structural Representations for Molecules and Materials
Musil, F., Grisafi, A., Bartók, A., Ortner, C., Cśanyi, G., Ceriotti, M., et al. (2021) · Cambridge University Engineering Department Publications Database · Cited by 440
Self-citationBottom-up Coarse-Graining: Principles and Perspectives
Jin, J., Pak, A. J., Durumeric, A. E. P., Loose, T. D., & Voth, G. A. (2022) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.2c00643 · Cited by 321
Machine learning approaches for the prediction of materials properties
Chibani, S., & Coudert, F. (2020) · APL Materials · doi:10.1063/5.0018384 · Cited by 273
Amici, J., Asinari, P., Ayerbe, E., Barboux, P., Bayle‐Guillemaud, P., Behm, R. J., Berecibar, M., Berg, E. J., Bhowmik, A., Bodoardo, S., et al. (2022) · Advanced Energy Materials · doi:10.1002/aenm.202102785 · Cited by 250
Perspective on integrating machine learning into computational chemistry and materials science
Westermayr, J., Gastegger, M., Schütt, K. T., & Maurer, R. J. (2021) · The Journal of Chemical Physics · doi:10.1063/5.0047760 · Cited by 216
How Machine Learning Will Revolutionize Electrochemical Sciences
Mistry, A. N., Franco, A. A., Cooper, S. J., Roberts, S. A., & Viswanathan, V. (2021) · ACS Energy Letters · doi:10.1021/acsenergylett.1c00194 · Cited by 197
Deep Learning for Deep Chemistry: Optimizing the Prediction of Chemical Patterns
Cova, T. F. G. G., & Pais, A. A. C. C. (2019) · Frontiers in Chemistry · doi:10.3389/fchem.2019.00809 · Cited by 193
Taking the leap between analytical chemistry and artificial intelligence: A tutorial review
Ayres, L. B., Gómez, F. J. V., Linton, J. R., Silva, M. F., & García, C. D. (2021) · Analytica Chimica Acta · doi:10.1016/j.aca.2021.338403 · Cited by 184
Material machine learning for alloys: Applications, challenges and perspectives
Liu, X., Xu, P., Zhao, J., Lu, W., Li, M., & Wang, G. (2022) · Journal of Alloys and Compounds · doi:10.1016/j.jallcom.2022.165984 · Cited by 157
Human- and machine-centred designs of molecules and materials for sustainability and decarbonization
Peng, J., Schwalbe‐Koda, D., Akkiraju, K., Xie, T., Giordano, L., Yu, Y., Eom, C. J., Lunger, J. R., Zheng, D. J., Rao, R. R., et al. (2022) · Nature Reviews Materials · doi:10.1038/s41578-022-00466-5 · Cited by 157
Neural network potentials for chemistry: concepts, applications and prospects
Käser, S., Vazquez-Salazar, L. I., Meuwly, M., & Töpfer, K. (2022) · Digital Discovery · doi:10.1039/d2dd00102k · Cited by 133
Mapping Materials and Molecules
Cheng, B., Griffiths, R., Wengert, S., Künkel, C., Stenczel, T. K., Zhu, B., Deringer, V. L., Bernstein, N., Margraf, J. T., Reuter, K., et al. (2020) · Accounts of Chemical Research · doi:10.1021/acs.accounts.0c00403 · Cited by 130
Machine learning for condensed matter physics
Bedolla-Montiel, E. A., Padierna, L. C., & Castañeda-Priego, R. (2020) · Journal of Physics Condensed Matter · doi:10.1088/1361-648x/abb895 · Cited by 134
A bright future for engineering piezoelectric 2D crystals
Sherrell, P. C., Fronzi, M., Shepelin, N. A., Corletto, A., Winkler, D. A., Ford, M. J., Shapter, J. G., & Ellis, A. V. (2021) · Chemical Society Reviews · doi:10.1039/d1cs00844g · Cited by 120
First-principles study of alkali-metal intercalation in disordered carbon anode materials
Huang, J., Cśanyi, G., Zhao, J., Cheng, J. J., & Deringer, V. L. (2019) · Journal of Materials Chemistry A · doi:10.1039/c9ta05453g · Cited by 103
SchNetPack 2.0: A neural network toolbox for atomistic machine learning
Schütt, K. T., Hessmann, S. S. P., Gebauer, N. W. A., Lederer, J., & Gastegger, M. (2023) · The Journal of Chemical Physics · doi:10.1063/5.0138367 · Cited by 108
Using Gaussian process regression to simulate the vibrational Raman spectra of molecular crystals
Raimbault, N., Grisafi, A., Ceriotti, M., & Rossi, M. (2019) · New Journal of Physics · doi:10.1088/1367-2630/ab4509 · Cited by 87
Self-citationBonati, L., Trizio, E., Rizzi, A., & Parrinello, M. (2023) · The Journal of Chemical Physics · doi:10.1063/5.0156343 · Cited by 90
Zhai, C., Li, T., Shi, H., & Yeo, J. (2020) · Journal of Materials Chemistry B · doi:10.1039/d0tb00896f · Cited by 86
An accurate machine-learning calculator for optimization of Li-ion battery cathodes
Houchins, G., & Viswanathan, V. (2020) · The Journal of Chemical Physics · doi:10.1063/5.0015872 · Cited by 82
In Silico Chemical Experiments in the Age of AI: From Quantum Chemistry to Machine Learning and Back
Aldossary, A., Campos-Gonzalez-Angulo, J. A., Pablo‐García, S., Leong, S. X., Rajaonson, E. M., Thiede, L., Tom, G., Wang, A., Avagliano, D., & Aspuru‐Guzik, A. (2024) · Advanced Materials · doi:10.1002/adma.202402369 · Cited by 92
Machine Learning-Guided Approach for Studying Solvation Environments
Basdogan, Y., Groenenboom, M. C., Henderson, E., De, S., Rempe, S. B., & Keith, J. A. (2019) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.9b00605 · Cited by 70
Switching of Redox Levels Leads to High Reductive Stability in Water-in-Salt Electrolytes
Wang, F., Sun, Y., & Cheng, J. J. (2023) · Journal of the American Chemical Society · doi:10.1021/jacs.2c11793 · Cited by 81
Machine Learning for the Discovery, Design, and Engineering of Materials.
Duan, C., Nandy, A., & Kulik, H. J. (2022) · Annual Review of Chemical and Biomolecular Engineering · doi:10.1146/annurev-chembioeng-092320-120230 · Cited by 69
Ramanathan, A., Ma, H., Parvatikar, A., & Chennubhotla, C. S. (2021) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2020.12.001 · Cited by 57
Prediction and Construction of Energetic Materials Based on Machine Learning Methods
Zang, X., Zhou, X., Bian, H., Jin, W., Pan, X., Jiang, J., Koroleva, M. Y., & Shen, R. (2022) · Molecules · doi:10.3390/molecules28010322 · Cited by 57
Identifying and Tracking Defects in Dynamic Supramolecular Polymers
Gasparotto, P., Bochicchio, D., Ceriotti, M., & Pavan, G. M. (2019) · The Journal of Physical Chemistry B · doi:10.1021/acs.jpcb.9b11015 · Cited by 52
Self-citationEnhanced Sampling in the Age of Machine Learning: Algorithms and Applications
Zhu, K., Trizio, E., Zhang, J., Hu, R., Jiang, L., Hou, T., & Bonati, L. (2025) · Chemical Reviews · doi:10.1021/acs.chemrev.5c00700 · Cited by 65
Tetef, S., Govind, N., & Seidler, G. T. (2021) · Physical Chemistry Chemical Physics · doi:10.1039/d1cp02903g · Cited by 56
Artificial neural networks for predicting charge transfer coupling
Wang, C.-I., Joanito, I., Lan, C., & Hsu, C. (2020) · The Journal of Chemical Physics · doi:10.1063/5.0023697 · Cited by 50
Efficient implementation of atom-density representations
Musil, F., Veit, M., Goscinski, A., Fraux, G., Willatt, M. J., Stricker, M., Junge, T., & Ceriotti, M. (2021) · The Journal of Chemical Physics · doi:10.1063/5.0044689 · Cited by 50
Self-citationMaking machine learning a useful tool in the accelerated discovery of transition metal complexes
Kulik, H. J. (2019) · Wiley Interdisciplinary Reviews Computational Molecular Science · doi:10.1002/wcms.1439 · Cited by 49
Edström, K., Dominko, R., Fichtner, M., & Perraud, S. (2020) · doi:10.33063/diva2-1452023 · Cited by 46
Structure-property maps with Kernel principal covariates regression
Helfrecht, B. A., Cersonsky, R. K., Fraux, G., & Ceriotti, M. (2020) · Machine Learning Science and Technology · doi:10.1088/2632-2153/aba9ef · Cited by 43
Self-citationEfficient implementation of atom-density representations
Musil, F., Veit, M., Goscinski, A., Fraux, G., Willatt, M. J., Stricker, M., Junge, T., & Ceriotti, M. (2022) · Refubium (Universitätsbibliothek der Freien Universität Berlin) · doi:10.17169/refubium-34235 · Cited by 46
Self-citationA new kind of atlas of zeolite building blocks
Helfrecht, B. A., Semino, R., Pireddu, G., Auerbach, S. M., & Ceriotti, M. (2019) · The Journal of Chemical Physics · doi:10.1063/1.5119751 · Cited by 45
Self-citationData-Driven Molecular Dynamics: A Multifaceted Challenge
Bernetti, M., Bertazzo, M., & Masetti, M. (2020) · Pharmaceuticals · doi:10.3390/ph13090253 · Cited by 43
Reaction coordinates in complex systems-a perspective
Rogal, J. (2021) · The European Physical Journal B · doi:10.1140/epjb/s10051-021-00233-5 · Cited by 45
E-Commerce System for Sale Prediction Using Machine Learning Technique
Singh, K., Marikannan, B. P., & Eaganathan, U. (2020) · Journal of Physics Conference Series · doi:10.1088/1742-6596/1712/1/012042 · Cited by 45
Artificial intelligence interventions in 2D MXenes-based photocatalytic applications
Mahapatra, D. M., Kumar, A., Kumar, R., Gupta, N. K., Ethiraj, B., & Singh, L. (2025) · Coordination Chemistry Reviews · doi:10.1016/j.ccr.2025.216460 · Cited by 39
Litman, Y. E., Behler, J., & Rossi, M. (2019) · Faraday Discussions · doi:10.1039/c9fd00056a · Cited by 40
Rossi, K., Jurásková, V., Wischert, R., Garel, L., Corminbœuf, C., & Ceriotti, M. (2020) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.0c00362 · Cited by 39
Self-citationDriving and characterizing nucleation of urea and glycine polymorphs in water
Zou, Z., Beyerle, E. R., Tsai, S.-T., & Tiwary, P. (2023) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2216099120 · Cited by 38
Godeffroy, L., Lemineur, J., Shkirskiy, V., Vieira, M. M., Noël, J., & Kanoufi, F. (2022) · Small Methods · doi:10.1002/smtd.202200659 · Cited by 37
Classifying soft self-assembled materials via unsupervised machine learning of defects
Gardin, A., Perego, C., Doni, G., & Pavan, G. M. (2022) · Communications Chemistry · doi:10.1038/s42004-022-00699-z · Cited by 37
Duan, C., Liu, F., Nandy, A., & Kulik, H. J. (2020) · The Journal of Physical Chemistry Letters · doi:10.1021/acs.jpclett.0c02018 · Cited by 35
Offei-Danso, A., Hassanali, A. A., & Rodríguez, Á. (2022) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.1c01292 · Cited by 34
Multiscale molecular modelling: from electronic structure to dynamics of nanosystems and beyond
Salahub, D. R. (2022) · Physical Chemistry Chemical Physics · doi:10.1039/d1cp05928a · Cited by 32
Exploring structure-property relationships in magnesium dissolution modulators
Würger, T., Mei, D., Vaghefinazari, B., Winkler, D. A., Lamaka, S. V., Zheludkevich, M. L., Meißner, R. H., & Feiler, C. (2021) · npj Materials Degradation · doi:10.1038/s41529-020-00148-z · Cited by 30
A Comprehensive Survey on Knowledge-Defined Networking
Wijesekara, P. A. D. S. N., & Gunawardena, S. (2023) · Telecom · doi:10.3390/telecom4030025 · Cited by 33
Armaković, S. J., Armaković, S., Bilić, A., & Savanović, M. M. (2025) · Catalysts · doi:10.3390/catal15080793 · Cited by 32
Moses, O. A., Chen, W., Adam, M. L., Wang, Z., Liu, K., Shao, J., Li, Z., Li, W., Wang, C., Zhao, H., et al. (2021) · Materials Reports Energy · doi:10.1016/j.matre.2021.100049 · Cited by 29
Wang, F., & Cheng, J. J. (2022) · Chemical Science · doi:10.1039/d2sc04025e · Cited by 29
Uncertainty quantification for predictions of atomistic neural networks
Vazquez-Salazar, L. I., Boittier, E. D., & Meuwly, M. (2022) · Chemical Science · doi:10.1039/d2sc04056e · Cited by 30
Moulahoum, H., & Ghorbanizamani, F. (2024) · Food Chemistry · doi:10.1016/j.foodchem.2024.138800 · Cited by 28
Differentiable Molecular Simulations for Control and LearningPreprint
Wang, W., Axelrod, S., & Gómez‐Bombarelli, R. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.00868 · Cited by 44
Caruso, C., Cardellini, A., Crippa, M., Rapetti, D., & Pavan, G. M. (2023) · The Journal of Chemical Physics · doi:10.1063/5.0147025 · Cited by 27
Nicholas, T. C., Goodwin, A. L., & Deringer, V. L. (2020) · Chemical Science · doi:10.1039/d0sc03287e · Cited by 27
Machine learning-based structure–property predictions in silica aerogels
Abdusalamov, R., Pandit, P., Milow, B., Itskov, M., & Rege, A. (2021) · Soft Matter · doi:10.1039/d1sm00307k · Cited by 26
Duan, C., Chu, D. B. K., Nandy, A., & Kulik, H. J. (2022) · Chemical Science · doi:10.1039/d2sc00393g · Cited by 27
Desgranges, C., & Delhommelle, J. (2020) · Molecular Systems Design & Engineering · doi:10.1039/d0me00134a · Cited by 19
Rimola, A., Ferrero, S., Germain, A., Corno, M., & Ugliengo, P. (2020) · Minerals · doi:10.3390/min11010026 · Cited by 15
Tetef, S., Kashyap, V., Holden, W. M., Velian, A., Govind, N., & Seidler, G. T. (2022) · The Journal of Physical Chemistry A · doi:10.1021/acs.jpca.2c03635 · Cited by 23
Ligand additivity relationships enable efficient exploration of transition metal chemical space
Arunachalam, N., Gugler, S., Taylor, M. G., Duan, C., Nandy, A., Janet, J. P., Meyer, R., Oldenstaedt, J. A., Chu, D. B. K., & Kulik, H. J. (2022) · The Journal of Chemical Physics · doi:10.1063/5.0125700 · Cited by 19
Mapping fluid structuration to flow enhancement in nanofluidic channels
Agarwal, A., Arya, V., Golani, B., Bakli, C., & Chakraborty, S. K. (2023) · The Journal of Chemical Physics · doi:10.1063/5.0140765 · Cited by 22
Feng, J., Wang, H., Ji, Y., & Li, Y. (2021) · Nano Select · doi:10.1002/nano.202100006 · Cited by 21
Spectral Map: Embedding Slow Kinetics in Collective Variables
Rydzewski, J. (2023) · The Journal of Physical Chemistry Letters · doi:10.1021/acs.jpclett.3c01101 · Cited by 21
Machine learning of slow collective variables and enhanced sampling via spatial techniques
Gökdemir, T., & Rydzewski, J. (2025) · Chemical Physics Reviews · doi:10.1063/5.0245177 · Cited by 20
Application of Optimization Algorithms in Clusters
Srivastava, R. (2021) · Frontiers in Chemistry · doi:10.3389/fchem.2021.637286 · Cited by 20
Gaigeot, M. (2021) · Spectrochimica Acta Part A Molecular and Biomolecular Spectroscopy · doi:10.1016/j.saa.2021.119864 · Cited by 21
Manifold learning in atomistic simulations: a conceptual review
Rydzewski, J., Chen, M., & Valsson, Ó. (2023) · Machine Learning Science and Technology · doi:10.1088/2632-2153/ace81a · Cited by 24
Valero, M. C., & Raybaud, P. (2020) · Journal of Catalysis · doi:10.1016/j.jcat.2020.09.006 · Cited by 19
Fanourgakis, G. S., Gkagkas, K., & Froudakis, G. E. (2022) · The Journal of Chemical Physics · doi:10.1063/5.0075994 · Cited by 21
Unsupervised topological learning for identification of atomic structures
Becker, S., Devijver, É., Molinier, R., & Jakse, N. (2022) · Physical review. E · doi:10.1103/physreve.105.045304 · Cited by 1
Duan, C., Ladera, A., Liu, J. C. L., Taylor, M. G., Ariyarathna, I. R., & Kulik, H. J. (2022) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.2c00468 · Cited by 18
Computational Design of Peptides for Biomaterials Applications
Wang, Y., Stebe, K. J., de la Fuente‐Núñez, C., & Radhakrishnan, R. (2023) · ACS Applied Bio Materials · doi:10.1021/acsabm.2c01023 · Cited by 19
Unsupervised topological learning approach of crystal nucleation
Becker, S., Devijver, É., Molinier, R., & Jakse, N. (2022) · Scientific Reports · doi:10.1038/s41598-022-06963-5 · Cited by 18
Unsupervised identification of crystal defects from atomistic potential descriptors
Kývala, L., de Hijes, P. M., & Christoph Dellago, A. (2025) · npj Computational Materials · doi:10.1038/s41524-025-01544-2 · Cited by 17
Wasserstein metric for improved quantum machine learning with adjacency matrix representations
Çaylak, O., von Lilienfeld, O. A., & Baumeier, B. (2020) · Machine Learning Science and Technology · doi:10.1088/2632-2153/aba048 · Cited by 16
Visualizing the Residue Interaction Landscape of Proteins by Temporal Network Embedding
Franke, L., & Peter, C. (2023) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.2c01228 · Cited by 18
Aquistapace, F., Amigó, N., Troncoso, J. F., Deluigi, O. R., & Bringa, E. M. (2023) · Computational Materials Science · doi:10.1016/j.commatsci.2023.112263 · Cited by 17
Machine learning a molecular Hamiltonian for predicting electron dynamics
Bhat, H. S., Ranka, K., & Isborn, C. M. (2020) · International Journal of Dynamics and Control · doi:10.1007/s40435-020-00699-8 · Cited by 16
Bougueroua, S., Bricage, M., Aboulfath, Y., Barth, D., & Gaigeot, M. (2023) · Molecules · doi:10.3390/molecules28072892 · Cited by 17
Understanding Defects in Amorphous Silicon with Million‐Atom Simulations and Machine Learning
Morrow, J. D., Ugwumadu, C. E., Drabold, D. A., Elliott, S. R., Goodwin, A. L., & Deringer, V. L. (2024) · Angewandte Chemie International Edition · doi:10.1002/anie.202403842 · Cited by 15
Shi, W., Jia, T., & Li, A. (2020) · Physical Chemistry Chemical Physics · doi:10.1039/d0cp01941k · Cited by 16
Classification of clusters in collision cascades
Bhardwaj, U., Sand, A. E., & Warrier, M. K. (2019) · Computational Materials Science · doi:10.1016/j.commatsci.2019.109364 · Cited by 15
Lewis, A. M., Lazzaroni, P., & Rossi, M. (2023) · The Journal of Chemical Physics · doi:10.1063/5.0154710 · Cited by 15
Element similarity in high-dimensional materials representations
Onwuli, A., Hegde, A. V., Nguyen, K. V. T., Butler, K. T., & Walsh, A. (2023) · Digital Discovery · doi:10.1039/d3dd00121k · Cited by 15
Unsupervised Machine Learning in the Analysis of Nonadiabatic Molecular Dynamics Simulation
Zhu, Y., Peng, J., Xu, C., & Lan, Z. (2024) · The Journal of Physical Chemistry Letters · doi:10.1021/acs.jpclett.4c01751 · Cited by 16
Ranking the synthesizability of hypothetical zeolites with the sorting hat
Helfrecht, B. A., Pireddu, G., Semino, R., Auerbach, S. M., & Ceriotti, M. (2022) · Digital Discovery · doi:10.1039/d2dd00056c · Cited by 15
Self-citationMean-field density matrix decompositions
Eriksen, J. J. (2020) · The Journal of Chemical Physics · doi:10.1063/5.0030764 · Cited by 17
Machine learning-driven insights into the microstructure and properties of high-entropy alloys
Zhang, X., Zhou, W., Li, X., Xu, T., Yu, Y., Zheng, L., Jin, G., & Zhang, S. (2025) · Advanced Powder Materials · doi:10.1016/j.apmate.2025.100331 · Cited by 15
Coarse-Grained Artificial Intelligence for Design of Brush Networks
Dobrynin, A. V., Stroujkova, A. F., Vatankhah‐Varnosfaderani, M., & Sheiko, S. S. (2023) · ACS Macro Letters · doi:10.1021/acsmacrolett.3c00479 · Cited by 16
Zhu, Y., Peng, J., Kang, X., Xu, C., & Lan, Z. (2022) · Physical Chemistry Chemical Physics · doi:10.1039/d2cp03323b · Cited by 13
Fabrizio, A., Meyer, B., & Corminbœuf, C. (2020) · The Journal of Chemical Physics · doi:10.1063/5.0005039 · Cited by 13
Metric learning for kernel ridge regression: assessment of molecular similarity
Fabregat, R., van Gerwen, P., Haeberle, M., Eisenbrand, F., & Corminbœuf, C. (2022) · Machine Learning Science and Technology · doi:10.1088/2632-2153/ac8e4f · Cited by 12
Metadata
- Authors
- Ceriotti, M.
- Affiliations
- École Polytechnique Fédérale de Lausanne (CH)
- Corresponding author
- Ceriotti, Michele
- ORCID
- Ceriotti, Michele
- Venue
- The Journal of Chemical Physics
- Publisher
- AIP Publishing
- Published
- 2019-04-19
- Volume
- 150
- Issue
- 15
- Pages
- 150901-150901
- ISSN
- 0021-9606 (ISSN-L)
- 1089-7690
- OpenAlex
- W2924564326
- Semantic Scholar
- 1db247fc224065b15b85c26dcb31a91eb3b0b05f
- Cited by
- 173
- References
- 110
- Citations by year
- 2026: 8 · 2025: 24 · 2024: 18 · 2023: 24 · 2022: 32 · 2021: 25 · 2020: 28 · 2019: 14
- FWCI
- 8.93
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- The Journal of Chemical Physics (publishedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion, cc-by-nc-nd)
- arXiv (Cornell University)
- Funders
- H2020 European Research Council (677013-HBMAP)
- Topics
- Physical Sciences › Materials Science › Materials Chemistry › Machine Learning in Materials Science
- Physical Sciences › Chemistry › Physical and Theoretical Chemistry › Advanced Physical and Chemical Molecular Interactions
- Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Model Reduction and Neural Networks
- Keywords
- unsupervised learning, atomistic simulation, molecular dynamics, clustering, dimensionality reduction, coarse-graining, structure-property relationships, machine learning bias, structure representation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-08, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-08 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
