Pith. sign in

REVIEW 4 major objections 5 minor 47 references

This paper shows that formation-energy prediction learned from ordered perovskites transfers directly to high-entropy perovskite oxides, while HOMO-LUMO gap prediction does not and requires a small HEPO-specific training set.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-03 05:31 UTC pith:VTXEJWJM

load-bearing objection Useful property-dependent transfer benchmark for perovskite GNNs, but the core numbers rest on unpublished SQS/DFT data and there are no error bars; deserves a serious referee with data release and multi-seed runs. the 4 major comments →

arxiv 2607.29510 v1 pith:VTXEJWJM submitted 2026-07-31 cond-mat.mtrl-sci cond-mat.othercs.LG

Ordered-to-disordered transfer learning with graph neural networks for formation-energy and HOMO-LUMO gap prediction in high-entropy perovskite oxides

Panupol Untarabut , Narjes Jomaa , Sylvian Cadars , Olivier Masson , Samuel Bernard , Assil Bouzid , Santanu Saha This is my paper
classification cond-mat.mtrl-sci cond-mat.othercs.LG
keywords high-entropy perovskite oxidestransfer learninggraph neural networksformation energyHOMO-LUMO gapspecial quasirandom structuresALIGNNchemical disorder
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper asks whether machine-learning models trained on chemically ordered perovskite oxides can be reused for high-entropy perovskite oxides (HEPOs), where cation disorder makes density-functional-theory calculations expensive. Using a curated dataset of over 12,000 relaxed structures spanning ABO3, A2BB'O6, and SQS-generated HEPO perovskites, the authors show that formation-energy prediction transfers almost unchanged from the ordered to the disordered domain (blind HEPO error 1.98 meV/atom versus 2.47 meV/atom on ordered test structures), while HOMO-LUMO gap prediction degrades roughly tenfold (0.28 eV versus 0.03 eV) and recovers only when a small HEPO-specific training fraction is added. The authors attribute this asymmetry to the gap's sensitivity to local chemical environments and disorder, versus stability's dependence on more transferable bond-length and octahedral-tilting patterns. The practical conclusion is a screening recipe: angle-aware graph neural networks trained on ordered perovskites are ready for HEPO stability screening, but electronic-property screening needs a small HEPO calibration set.

Core claim

The central discovery is that ordered-to-disordered transfer learning in perovskite oxides is property-dependent: formation energies learned from ordered single and double perovskites transfer directly to disordered high-entropy perovskites (ALIGNN blind HEPO MAE 1.98 meV/atom, similar to its 2.47 meV/atom ordered test error), whereas HOMO-LUMO gap prediction shows a roughly tenfold error increase (0.28 eV versus 0.03 eV) with systematic underestimation. The paper further finds that this transfer deficit is largely correctable: adding only 20% of the HEPO training subset (16% of all HEPO structures) reduces the gap MAE to 0.07 eV, and using the full HEPO training set reaches 0.04 eV. Among f

What carries the argument

The central mechanism is the graph neural network representation of crystal structure, compared across four architectures: CGCNN (pairwise atom-bonds), GATGNN (attention-weighted graphs), ALIGNN (line-graph encoding of bond angles), and M3GNet (three-body geometric features). The load-bearing object is ALIGNN's explicit angular message passing, which allows the model to encode B-O-B angles and octahedral tilting—structural motifs previously shown to control perovskite stability and electronic structure. The transfer protocol itself is the other key piece: training only on chemically ordered ABO3/A2BB'O6 perovskites and evaluating on SQS-generated HEPO structures shares the same elemental poo

Load-bearing premise

The benchmark rests on the authors' own SQS-generated HEPO dataset (1810 structures) and DFT labels from their previous works, several of which are still 'to be submitted'; if those SQS models do not faithfully represent real HEPO disorder, or if the reference phases are inconsistent between the ordered and disordered domains, the transfer comparison collapses.

What would settle it

Regenerate HEPO structures with an independent disorder sampling method (e.g., a different SQS implementation, random supercells, or experimentally determined cation arrangements), recompute the same DFT targets, and re-run the blind transfer test; if formation-energy MAE rises well above ~2 meV/atom or HOMO-LUMO gap MAE falls below ~0.1 eV, the property-dependent transfer conclusion would not hold.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Formation-energy models trained on ordered perovskites can be used directly to screen HEPO thermodynamic stability without additional HEPO DFT data.
  • For HEPO electronic-structure screening, a small HEPO-specific calibration set (20% of a training split) is enough to cut the HOMO-LUMO gap error from 0.28 eV to 0.07 eV.
  • Angle-aware graph neural networks (ALIGNN) should be preferred over pairwise-only GNNs for perovskite property prediction, because angular encoding improves both accuracy and latent-space organization.
  • Part of the apparent transfer failure for HOMO-LUMO gaps is a validation-domain mismatch: using HEPO validation data for model selection halves the blind error from 0.28 eV to 0.14 eV even without HEPO training structures.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The property-dependent transfer asymmetry likely extends to other local-environment-sensitive electronic properties (band edges, effective masses, optical spectra), so the same ordered-to-disordered protocol could be used to decide where calibration data are needed.
  • The transfer success for formation energy suggests that bond-length and tilting descriptors are largely disorder-invariant for these chemistries; if true, a descriptor-based model (not necessarily a GNN) might achieve the same transfer with far less training data.
  • The SQS representation, by construction, captures only a finite set of configurations; the transfer conclusions should be tested against larger supercells or experimentally resolved structures before being used for quantitative HEPO discovery.
  • The observed non-monotonic formation-energy error with HEPO training fraction (minimum at 0.8) hints that adding disordered data can slightly disturb ordered-domain knowledge, suggesting a possible role for regularization or multi-task learning in future pipelines.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. This paper benchmarks four graph neural networks (CGCNN, GATGNN, ALIGNN, M3GNet) for predicting formation energy (Ef) and HOMO–LUMO gap (Eg) on a curated DFT dataset of ordered ABO3/A2BB'O6 perovskites (10,898 structures) and SQS-modeled high-entropy perovskite oxides (1,810 structures). The authors evaluate each model on held-out ordered perovskites, then blind-transfer the ordered-trained models to HEPO test structures, and finally fine-tune with fractions of HEPO training data. The central claims are: (i) ALIGNN is the most accurate model on the ordered domain; (ii) formation-energy prediction transfers effectively to HEPOs (blind MAE 1.98 meV/atom, close to the ordered test MAE of 2.47 meV/atom); (iii) HOMO–LUMO-gap prediction transfers poorly (blind HEPO MAE 0.28 eV vs 0.03 eV on ordered); and (iv) adding a small HEPO training fraction substantially improves gap prediction. UMAP embedding analysis is used to argue that angle-aware representations, especially ALIGNN, better organize chemistry and octahedral-tilting information.

Significance. If the empirical findings are robust, this is a useful practical contribution to ML-based screening of high-entropy perovskite oxides. The controlled design — fixed A/B elemental pools between ordered and disordered domains, composition-stratified splits, and a clearly described blind-transfer protocol — is a strength. The data-efficiency result, namely that a small HEPO-specific calibration set can close most of the gap-transfer error, is actionable. However, the central comparison currently depends on an unpublished SQS dataset with no released structures or external validation, and the paper lacks uncertainty quantification and trivial baselines. These gaps need to be addressed before the practical conclusions can be considered quantitatively established.

major comments (4)
  1. [Training models on ordered domain / Fig. 3] No uncertainty quantification is provided. Claims such as 'ALIGNN gives the best overall performance' rest on MAE differences as small as 0.01 eV for Eg (0.03 vs 0.04 eV) and, in the fine-tuning curve of Fig. 9(c), differences of 0.01 eV between fractions (e.g., 0.06 vs 0.05 eV). With a single training run per model and no seed variation, these differences are within typical run-to-run noise. Please report mean±std over multiple random seeds and, where relevant, paired statistical tests.
  2. [Transferability from ordered perovskites to HEPOs / Table 1, Fig. 5, Fig. 9] Trivial baselines are missing. The HEPO formation-energy test distributions are very narrow for Sr and Ba (Table 1: σ = 4.10 and 2.44 meV/atom). For a normal distribution, a family-mean predictor gives MAE ≈ 0.798σ, i.e., ≈3.3 and ≈1.9 meV/atom for Sr and Ba, respectively. The reported Ba HEPO MAE of 2.24 meV/atom (Fig. 5c) is actually worse than this baseline, and the overall blind MAE of 1.98 meV/atom is only modestly better than a composition-resolved mean baseline. Without such baselines on the HEPO test set, 'formation-energy transfer works' is not quantitatively established. Please add trivial baselines (e.g., composition family mean, composition plus A-site mean) and discuss the results relative to them.
  3. [Dataset construction / Methods / refs [33,34]] The entire transfer benchmark rests on the 1810 SQS-generated HEPO structures and their DFT labels, which are taken from the authors' own 'to be submitted' works (refs 33–34). No SQS structures, computed properties, or pseudopotential/reference-phase consistency details are released, and no independent external validation is provided. If the SQS models do not faithfully represent real HEPO disorder, or if the DFT reference phases/settings differ between the ordered and disordered domains, the central comparison could be an artifact. Please release the dataset or provide an independent reproduction, and explicitly document the consistency of DFT settings and reference phases across the ordered and HEPO domains.
  4. [Fig. 9(c), 'Transferability from ordered perovskites to HEPOs'] The 0.0 HEPO-fraction point is not a blind-transfer point: although no HEPO structures are in the training set, the HEPO validation set is used for model selection. The paper clearly acknowledges this, but the figure and surrounding text should make the distinction visually and verbally explicit. As presented, a reader may misinterpret the 0.14 eV value at fraction 0.0 as a blind-transfer error, when the true blind error is 0.28 eV. This does not invalidate the fine-tuning trend, but the labeling is important for the paper's central message.
minor comments (5)
  1. [Dataset construction] Typo: 'using a a 80:10:10 split' should read 'using an 80:10:10 split'.
  2. [Error analysis across chemical species] Inconsistent capitalization: 'M3GNET' appears where 'M3GNet' is used elsewhere.
  3. [Training behavior / Fig. 2] The M3GNet formation-energy training loss (2.12×10^-2) exceeds its validation loss (7.08×10^-3), which is unusual and may indicate a normalization or logging issue. Please clarify or verify.
  4. [Transferability / Fig. 9 caption] The caption of Fig. 9 should explicitly state that the '0.0' fraction uses HEPO validation for model selection, while the blind-transfer parity plots do not. This distinction appears only in the main text.
  5. [Introduction / References] Several key references (refs 31–34) are 'to be submitted' preprints; if the companion papers become available, the authors should update these references and ideally cite the actual datasets.

Circularity Check

0 steps flagged

No construction-level circularity: the transfer results are genuine held-out ML evaluations. Score 2 reflects the heavy reliance on the authors' own unpublished DFT/SQS datasets (refs 31-34), which limits independent verification but does not make the derivation circular.

full rationale

The paper's central claims are empirical benchmark results: GNNs are trained on ordered perovskites and evaluated on a never-seen HEPO test subset (80:10:10 split per A-site family). No equation in the paper defines a target in terms of the model output, no fitted parameter is relabeled as a prediction, and no uniqueness theorem is imported to force the chosen architecture. ALIGNN is selected on ordered-domain performance, and the HEPO fine-tuning uses the fixed HEPO test set only for final evaluation, with the HEPO validation set for model selection. The main dependency is that the HEPO structures and their DFT formation energies/gaps, as well as the ordered structures, are taken from the authors' previous works (refs 31-34), several marked 'to be submitted'. This is a reproducibility/evidence limitation, not a circular step: the ML predictions are not defined in terms of the DFT labels, and the transfer comparison is a legitimate held-out experiment on that dataset. A missing family-mean baseline for the low-variance Sr/Ba formation energies is a correctness risk, but it does not make the derivation circular.

Axiom & Free-Parameter Ledger

0 free parameters · 4 axioms · 0 invented entities

No new physical entities or free parameters are introduced. The load-bearing inputs are the DFT dataset and SQS models from the authors' own previous works, the exactness of which is assumed. Free parameters: none beyond ordinary ML training; the model hyperparameters are chosen by validation and reported in the SI.

axioms (4)
  • domain assumption DFT-PBE labels (formation energies and HOMO-LUMO gaps) from the authors' previous works are accurate and mutually consistent
    All target values in the ordered and HEPO datasets are computed in prior papers (refs 31-34) and taken as ground truth; the entire transfer benchmark inherits any systematic DFT error or reference-phase inconsistency (Section 'Dataset construction').
  • domain assumption SQS supercells faithfully represent the chemical/structural disorder of HEPOs
    The HEPO domain is represented only by special quasirandom structures (refs 33-35); if SQS does not capture the relevant configurational diversity, transfer conclusions on these 1810 structures may not hold for real HEPOs (Section 'Dataset construction', 'Transferability from ordered perovskites to HEPOs').
  • domain assumption Element pool restriction to A = [Ca,Sr,Ba] and B = [Ti,Zr,Hf,Sn,Ge] supports generalizable conclusions
    The paper explicitly restricts all structures to 3 A-site and 5 B-site elements; transfer performance on this limited chemistry may not extend to other cations or oxidation states (Section 'Dataset construction').
  • domain assumption UMAP projections of graph embeddings are informative evidence for what the models learned
    The embedding analysis uses UMAP after PCA and is interpreted as showing organization by chemistry and tilting (Section 'Feature embedding and UMAP analysis'); this is a qualitative visualization assumption, not a rigorous structural probe.

pith-pipeline@v1.3.0-daily-deepseek · 14210 in / 6996 out tokens · 68796 ms · 2026-08-03T05:31:35.528943+00:00 · methodology

0 comments
read the original abstract

High-entropy perovskite oxides (HEPOs) represent a chemically complex class of materials with promising functional properties, yet their vast compositional space and, chemical/structural disorder pose significant challenge for accurate property prediction. Graph neural networks (GNNs) enable rapid exploration of materials space but are often limited by the availability of representative training data. Here, we investigate ordered-to-disordered transfer learning using GNNs for formation-energy and HOMO-LUMO gap prediction in HEPOs by transferring knowledge learned from chemically ordered perovskites. Four representative GNN models, including CGCNN, GATGNN, ALIGNN and M3GNet are evaluated to understand the role of structural representations, spanning pairwise two-body and angular three-body interactions in transfer performance. We find strong property-dependent transfer behavior: formation-energy prediction transfers effectively to disordered HEPOs, whereas HOMO-LUMO gap prediction shows limited transferability due to its sensitivity to local chemical environments. Incorporating a small HEPO-specific training dataset substantially improves HOMO-LUMO gap prediction. Representation-level analysis using UMAP further highlights the importance of encoding three-body geometric information such as in ALIGNN for capturing complex structure-property relationships and improving transferability.

Figures

Figures reproduced from arXiv: 2607.29510 by Assil Bouzid, Narjes Jomaa, Olivier Masson, Panupol Untarabut, Samuel Bernard, Santanu Saha, Sylvian Cadars.

Figure 1
Figure 1. Figure 1: Target-property distributions for the training, validation, and test subsets of the ordered single [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Training behavior of the graph neural network models on the training and validation set of [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Train, validation, and test MAE values for the ordered domain (single and double perovskite). (a) [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Prediction quality of the trained GNN models on the ordered single and double perovskite test set. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Chemistry-resolved prediction errors on held out test set of ordered and disordered domain. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: UMAP projections of graph-level embeddings for (a) CGCNN formation-energy prediction, (b) [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: UMAP projections of graph-level embeddings for (a) CGCNN formation-energy prediction and [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: HEPO target-property distributions across the training, validation, and test subsets. (a–c) DFT [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Ordered perovskite to HEPO transferability analysis. Blind-test parity plots comparing DFT refer [PITH_FULL_IMAGE:figures/full_fig_p013_9.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

47 extracted references · 3 linked inside Pith

  1. [1]

    Rost, C. M.et al. Entropy-stabilized oxides.Nat. Commun.6, 8485 (2015). 16

    2015

  2. [2]

    & Reece, M

    Zhang, R.-Z. & Reece, M. J. Review of high entropy ceramics: design, synthesis, structure and proper- ties. J. Mater. Chem. A7, 22148–22162 (2019)

    2019

  3. [3]

    A new class of high-entropy perovskite oxides.Scr

    Jiang, S.et al. A new class of high-entropy perovskite oxides.Scr. Mater.142, 116–120 (2018)

    2018

  4. [4]

    Sarkar, A. et al. Rare earth and transition metal based entropy stabilised perovskite type oxides. J. Eur. Ceram. Soc.38, 2318–2327 (2018)

    2018

  5. [5]

    Shah, S. A.et al. First-principles calculations to investigate the structural, electronic, optical, mechan- ical, and thermodynamic properties of double perovskites Ba2WB′O6 (B′ = co, fe, mn, ni, and zn). Optik300, 171636 (2024)

    2024

  6. [6]

    High-entropy materials for energy and electronic applications.Nat

    Schweidler, S.et al. High-entropy materials for energy and electronic applications.Nat. Rev. Mater.9, 266–281 (2024)

    2024

  7. [7]

    Almishal, S. S.et al. Thermodynamics-inspired high-entropy oxide synthesis.Nat. Commun.16, 8211 (2025)

    2025

  8. [8]

    Su, L. et al. Direct observation of elemental fluctuation and oxygen octahedral distortion-dependent charge distribution in high entropy oxides.Nat. Commun.13, 2358 (2022)

    2022

  9. [9]

    & Wang, D

    Liang, Y., Luo, B., Dong, H. & Wang, D. Electronic structure and transport properties of sol-gel-derived high-entropy ba (zr0. 2sn0. 2ti0. 2hf0. 2nb0. 2) o3 thin films.Ceram. Int.47, 20196–20200 (2021)

    2021

  10. [10]

    & Wang, J

    Bai, Z., Luo, B., Peng, T. & Wang, J. High-entropy perovskite oxide photonic synapses. Adv. Opt. Mater.12, 2303248 (2024)

    2024

  11. [11]

    & Kohn, W

    Hohenberg, P. & Kohn, W. Inhomogeneous electron gas.Phys. Rev.136, B864–B871 (1964)

    1964

  12. [12]

    & Sham, L

    Kohn, W. & Sham, L. J. Self-consistent equations including exchange and correlation effects.Phys. Rev. 140, A1133–A1138 (1965)

    1965

  13. [13]

    Emery, A. A. & Wolverton, C. High-throughput DFT calculations of formation energy, stability and oxygen vacancy formation energy of ABO3 perovskites. Sci. Data4, 170153 (2017)

    2017

  14. [14]

    J., Morelock, R

    Bare, Z. J., Morelock, R. J. & Musgrave, C. B. Dataset of theoretical multinary perovskite oxides. Sci. Data10, 244 (2023)

    2023

  15. [15]

    Kim, J.-S., Noh, J. & Im, J. Machine learning-enabled chemical space exploration of all-inorganic perovskites for photovoltaics.npj Comp. Mater.10, 97 (2024)

    2024

  16. [16]

    Predicting the formation of fractionally doped perovskite oxides by a function-confined machine learning method.Commun

    Zhai, X.et al. Predicting the formation of fractionally doped perovskite oxides by a function-confined machine learning method.Commun. Mater.3, 42 (2022)

    2022

  17. [17]

    Transfer learning guided discovery of efficient perovskite oxide for alkaline water oxida- tion

    Jiang, C.et al. Transfer learning guided discovery of efficient perovskite oxide for alkaline water oxida- tion. Nat. Commun.15, 6301 (2024)

    2024

  18. [18]

    Moon, J. et al. Active learning guides discovery of a champion four-metal perovskite oxide for oxygen evolution electrocatalysis. Nat. Mater.23, 108–115 (2024)

    2024

  19. [19]

    & Liu, Y

    Li, Y., Zhu, R., Wang, Y., Feng, L. & Liu, Y. Center-environment deep transfer machine learning across crystal structures: from spinel oxides to perovskite oxides.npj Comp. Mater.9, 109 (2023)

    2023

  20. [20]

    Gupta, V. et al. Structure-aware graph neural network based deep transfer learning framework for enhanced predictive analytics on diverse materials datasets.npj Comp. Mater.10, 1 (2024)

    2024

  21. [21]

    & Karppinen, M.A2BB ′O6 perovskites: A review.Prog

    Vasala, S. & Karppinen, M.A2BB ′O6 perovskites: A review.Prog. Solid State Chem.43, 1–36 (2015)

    2015

  22. [22]

    Chen, X., Xu, J., Xu, Y., Luo, F. & Du, Y. Rare earth double perovskites: A fertile soil in the field of perovskite oxides. Inorg. Chem. Front.6, 2226–2238 (2019). 17

    2019

  23. [23]

    Glazer, A. M. The classification of tilted octahedra in perovskites.Acta Cryst. B28, 3384–3392 (1972)

    1972

  24. [24]

    Woodward, P. M. Octahedral tilting in perovskites. i. geometrical considerations.Acta Cryst. B53, 32–43 (1997)

    1997

  25. [25]

    Effects of octahedral tilting on the electronic structure and optical properties ofd0 double perovskitesA 2ScSbO6 (A= Sr, Ca)

    Ray, R.et al. Effects of octahedral tilting on the electronic structure and optical properties ofd0 double perovskitesA 2ScSbO6 (A= Sr, Ca). J. Alloys Compd.705, 497–506 (2017)

    2017

  26. [26]

    P., Shaikh, M

    Ghosh, A., Palanichamy, G., Trujillo, D. P., Shaikh, M. & Ghosh, S. Insights into cation ordering of double perovskite oxides from machine learning and causal relations.Chem. Mater.34, 7563–7578 (2022)

    2022

  27. [27]

    & Grossman, J

    Xie, T. & Grossman, J. C. Crystal graph convolutional neural networks for an accurate and interpretable prediction of material properties.Phys. Rev. Lett.120, 145301 (2018)

    2018

  28. [28]

    Louis, S.-Y. et al. Graph convolutional neural networks with global attention for improved materials property prediction. Phys. Chem. Chem. Phys.22, 18141–18148 (2020)

    2020

  29. [29]

    & DeCost, B

    Choudhary, K. & DeCost, B. Atomistic line graph neural network for improved materials property predictions. npj Comp. Mater.7, 185 (2021)

    2021

  30. [30]

    & Ong, S

    Chen, C. & Ong, S. P. A universal graph deep learning interatomic potential for the periodic table. Nat. Comp. Sci.2, 718–728 (2022)

    2022

  31. [31]

    Mapping the influence of symmetry breaking in structure-property relationships of ABO3 perovskites (to be submitted).arXiv preprint arXiv:2607.28025 (2026).2607.28025

    Untarabut, P.et al. Mapping the influence of symmetry breaking in structure-property relationships of ABO3 perovskites (to be submitted).arXiv preprint arXiv:2607.28025 (2026).2607.28025

    Pith/arXiv arXiv 2026

  32. [32]

    Interplay between cation ordering and octahedral tilting in double perovskite oxides A2BB ′O6: a first-principles bridge to disordered perovskite oxides (to be submitted) (2026)

    Untarabut, P.et al. Interplay between cation ordering and octahedral tilting in double perovskite oxides A2BB ′O6: a first-principles bridge to disordered perovskite oxides (to be submitted) (2026)

    2026

  33. [33]

    Jomaa, N. et al. An aiida-atat plugin for automated generation of special quasirandom structures: Application to high-entropy perovskite oxides (to be submitted) (2026)

    2026

  34. [34]

    Untarabut, P. et al. Ab-initio investigation of high-entropy perovskite oxide A(T i0.2Zr 0.2Hf 0.2Sn0.2Ge0.2)O3 for A=[Ca,Sr,Ba] (to be submitted) (2026)

    2026

  35. [35]

    Specialquasirandomstructures.Phys

    Zunger, A., Wei, S.-H., Ferreira, L.G.&Bernard, J.E. Specialquasirandomstructures.Phys. Rev. Lett. 65, 353–356 (1990)

    1990

  36. [36]

    Kingma, D. P. & Ba, J. Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980 (2014)

    Pith/arXiv arXiv 2014

  37. [37]

    J., Guennou, M., Iniguez, J., Kreisel, J

    Xiang, H. J., Guennou, M., Iniguez, J., Kreisel, J. & Bellaiche, L. Rules and mechanisms governing octahedral tilts in perovskites under pressure.Phys. Rev. B96, 054102 (2017)

    2017

  38. [38]

    Yan, K., Liu, Y., Lin, Y. & Ji, S. Periodic graph transformers for crystal material property prediction. Advances in Neural Information Processing Systems35, 15066–15080 (2022)

    2022

  39. [39]

    & Melville, J

    McInnes, L., Healy, J. & Melville, J. UMAP: Uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426 (2018).1802.03426

    Pith/arXiv arXiv 2018

  40. [40]

    Jolliffe, I. T. & Cadima, J. Principal component analysis: A review and recent developments. Phil. Trans. R. Soc. A374, 20150202 (2016)

    2016

  41. [41]

    QUANTUM ESPRESSO: A modular and open-source software project for quantum simulations of materials.J

    Giannozzi, P.et al. QUANTUM ESPRESSO: A modular and open-source software project for quantum simulations of materials.J. Phys. Condens. Matter21, 395502 (2009)

    2009

  42. [42]

    Giannozzi, P. et al. Advanced capabilities for materials modelling with Quantum ESPRESSO. J. Phys. Condens. Matter29, 465901 (2017)

    2017

  43. [43]

    P., Burke, K

    Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett.77, 3865–3868 (1996). 18

    1996

  44. [44]

    Hamann, D. R. Optimized norm-conserving Vanderbilt pseudopotentials. Phys. Rev. B88, 085117 (2013)

    2013

  45. [45]

    van Setten, M. J.et al. The PseudoDojo: Training and grading a 85 element optimized norm-conserving pseudopotential table. Comp. Phys. Commun.226, 39–54 (2018)

    2018

  46. [46]

    Lejaeghere, K. et al. Reproducibility in density functional theory calculations of solids.Science351, aad3000 (2016)

    2016

  47. [47]

    Monkhorst, H. J. & Pack, J. D. Special points for Brillouin-zone integrations.Phys. Rev. B13, 5188– 5192 (1976). 19

    1976