arXiv subjects

Danny Perez

Publications and source records attributed to Danny Perez.

At least 19 recordsLinked to original sources

arXiv2026

Surrogate-Gated Generation and Foundation-Model Embeddings for Bayesian Materials Design

Closed-loop materials discovery iterates between proposing candidate structures and evaluating their properties, and property evaluation dominates the cost. In the generative variant, a learned prior proposes candidate crystals and a property oracle scores them; we ask whether a cheap probabilistic surrogate can triage the generator's output, and what such a surrogate must do well. Across three architecturally distinct pretrained diffusion priors (MatterGen, CrystalFlow, ADiT) and two targets (room-temperature heat capacity and bulk modulus), we insert a Gaussian process acquisition gate between structure generation and the oracle in an RL-steered generative workflow. The gate matches or exceeds ungated fine-tuning of the generative model while capping oracle calls at a fixed per-cycle budget. Budget-matched ablations isolate the mechanism. At an identical four-call budget, ranking-based selection outperforms arbitrary selection, confirming that the gain comes from the surrogate's choice; the gate comes within $\sim$9\% of exhaustive oracle spending at roughly one-fifth of the calls. A density-functional-theory check of the bulk-modulus discoveries confirms the learned oracle to within 2.5\% on average and the surrogate's ranking of the generated structures at Spearman $\rho = 0.94$. A cross-factorial benchmark of surrogate performance spanning mechanical, electronic, and vibrational properties identifies pretrained ORB embeddings with a Gaussian process as the most reliable combination, which we adopt as the building blocks of the proposed workflow. The complete pipeline is released as open-source software.

cond-mat.mtrl-sci↗

arXiv2026

Characterizing Atomistic Transitions Using Cross-scale Graph-pooled Chebyshev Signatures

Large-scale atomistic simulations can produce extreme volumes of information in the form of long trajectories. Reliably and automatically extracting key information from such datasets remains a formidable challenge, especially as it pertains to the analysis of the structural transitions affecting the system. We present a novel approach to characterize and compare atomistic transitions using cross-scale graph-pooled Chebyshev signatures. These signatures are permutation invariants of an operator that transform a Coulomb matrix representation of the initial state of the system into that corresponding to the final state. Using a long-time trajectory of a small metallic nanoparticle, we show that these signatures can be used to define a natural distance metric between transitions that allows for classification and clustering into physically meaningful families. This approach is shown to capture complex patterns and hierarchies of transition types that are inaccessible to traditional techniques, dramatically facilitating the analysis of large-scale simulations.

physics.comp-ph↗

arXiv2026

Predicting Atomistic Transitions with Transformers

Accurate knowledge of the atomistic transition pathways in materials and material surfaces is crucial for many material science problems. However, conventional simulation techniques used to find these transitions are extremely computationally intensive. Even with large-scale, accelerated material simulations, the computational cost constrains the applicable domain in practice. Machine learning models, with the potential to learn the complex emergent behaviors governing atomistic transitions as a fast surrogate model, have great promise to predict transitions with a vastly reduced computational cost. Here, we demonstrate how transformers can be trained to predict atomistic transitions in nano-clusters. We show how we evaluate physical validity of the predictions and how a multitude of additional, different microstates can be generated by slightly varying the data provided to the model.

cond-mat.mtrl-sci↗

arXiv2026

Exploring the extremes: atomic basis for multi-elemental materials science under complex thermodynamic conditions

Modern materials science has historically been founded on combining restricted subsets of the periodic table, favoring high-purity, few-element systems. However, the demands of an emerging circular economy, together with the need to understand materials behavior under planetary and industrial extremes, increasingly require mastering Mendeleev materials - chemically and structurally complex systems that span large portions of the periodic table. In these regimes, current universal machine-learning interatomic potentials often fail, largely due to systematic gaps in traditional training datasets that heavily emphasize low-energy, near-equilibrium structures. We address this limitation by introducing a chemistry-agnostic, information-entropy-maximization protocol for data generation. By decoupling structural sampling from thermodynamic bias, our approach provides a robust physical prior for atomic interactions across the entire periodic table, including regimes far from equilibrium and under extreme conditions. Training a Graph Atomic Cluster Expansion (GRACE) model on the resulting statistically maximized entropy (SMAX) dataset yields markedly improved robustness across a range of stringent benchmarks. These include large-strain phase transformations in tin, defect evolution in tungsten-based alloys, and catalytic reaction barrier prediction. More broadly, our approach establishes a scalable and principled methodology for navigating the vast chemical and configurational space relevant to future materials design. It enables a paradigm of discovery by simulation in which unbiased sampling protocols autonomously resolve emergent structures in multi-elemental mixtures-such as systems containing the nine most abundant elements in the Earth's crust-without reliance on a priori chemical assumptions.

cond-mat.mtrl-sci↗

arXiv2026

LAMDA: Aiding Visual Exploration of Atomic Displacements in Molecular Dynamics Simulations

Contemporary materials science research is heavily conducted in silico, involving massive simulations of the atomic-scale evolution of materials. Cataloging basic patterns in the atomic displacements is key to understanding and predicting the evolution of physical properties. However, the combinatorial complexity of the space of possible transitions coupled with the overwhelming amount of data being produced by high-throughput simulations make such an analysis extremely challenging and time-consuming for domain experts. The development of visual analytics systems that facilitate the exploration of simulation data is an active field of research. While these systems excel in identifying temporal regions of interest, they treat each timestep of a simulation as an independent event without considering the behavior of the atomic displacements between timesteps. We address this gap by introducing LAMDA, a visual analytics system that allows domain experts to quickly and systematically explore state-to-state transitions. In LAMDA, transitions are hierarchically categorized, providing a basis for cataloging displacement behavior, as well as enabling the analysis of simulations at different resolutions, ranging from very broad qualitative classes of transitions to very narrow definitions of unit processes. LAMDA supports navigating the hierarchy of transitions, enabling scientists to visualize the commonalities between different transitions in each class in terms of invariant features characterizing local atomic environments, and LAMDA simplifies the analysis by capturing user inputs through annotations. We evaluate our system through a case study and report on findings from our domain experts.

cs.HC↗

arXiv2025

Deep Gaussian Process-based Cost-Aware Batch Bayesian Optimization for Complex Materials Design Campaigns

The accelerating pace and expanding scope of materials discovery demand optimization frameworks that efficiently navigate vast, nonlinear design spaces while judiciously allocating limited evaluation resources. We present a cost-aware, batch Bayesian optimization scheme powered by deep Gaussian process (DGP) surrogates and a heterotopic querying strategy. Our DGP surrogate, formed by stacking GP layers, models complex hierarchical relationships among high-dimensional compositional features and captures correlations across multiple target properties, propagating uncertainty through successive layers. We integrate evaluation cost into an upper-confidence-bound acquisition extension, which, together with heterotopic querying, proposes small batches of candidates in parallel, balancing exploration of under-characterized regions with exploitation of high-mean, low-variance predictions across correlated properties. Applied to refractory high-entropy alloys for high-temperature applications, our framework converges to optimal formulations in fewer iterations with cost-aware queries than conventional GP-based BO, highlighting the value of deep, uncertainty-aware, cost-sensitive strategies in materials campaigns.

cond-mat.mtrl-sci↗

arXiv2025

Active Learning of A Crystal Plasticity Flow Rule From Discrete Dislocation Dynamics Simulations

Continuum-scale material deformation models, such as crystal plasticity, can significantly enhance their predictive accuracy by incorporating input from lower-scale (i.e., mesoscale) models. The procedure to generate and extract the relevant information is however typically complex and ad hoc, involving decision and intervention by domain experts, leading to long development times. In this study, we develop a principled approach for calibration of continuum-scale models using lower scale information by representing a crystal plasticity flow rule as a Gaussian process model. This representation allows for efficient parameter space exploration, guided by the uncertainty embedded in the model through a process known as Bayesian optimization. We demonstrate a semi-autonomous Bayesian optimization loop which instantiates discrete dislocation dynamics simulations whose initial conditions are automatically chosen to optimize the uncertainty of a model crystal plasticity flow rule. Our self-guided computational pipeline efficiently generated a dataset and corresponding model whose error, uncertainty, and physical feature sensitivities were validated with comparison to an independent dataset four times larger, demonstrating a valuable and efficient active learning implementation readily transferable to similar material systems.

cond-mat.mtrl-sci↗

arXiv2025

Hop-Decorate: An Automated Atomistic Workflow for Generating Defect Transport Data in Chemically Complex Materials

Chemically complex materials (CCMs) exhibit extraordinary functional properties but pose significant challenges for atomistic modeling due to their vast configurational heterogeneity. We introduce Hop-Decorate (HopDec), a high-throughput, Python-based atomistic workflow that automates the generation of defect transport data in CCMs. HopDec integrates accelerated molecular dynamics with a novel redecoration algorithm to efficiently sample migration pathways across chemically diverse local environments. The method constructs a defect-state graph in which transitions are associated with distributions of kinetic and thermodynamic parameters, enabling direct input into kinetic Monte Carlo and other mesoscale models. We demonstrate HopDec's capabilities through applications to a Cu-Ni alloy and the spinel oxide (Fe,Ni)Cr2O4, revealing simple predictive relationships in the former and complex migration behaviors driven by cation disorder in the latter. These results highlight HopDec's ability to extract physically meaningful trends and support reduced-order or machine-learned models of defect kinetics, bridging atomic-scale simulations and mesoscale predictions in complex material systems.

cond-mat.mtrl-sci↗

arXiv2025

Accurate and Uncertainty-Aware Multi-Task Prediction of HEA Properties Using Prior-Guided Deep Gaussian Processes

Surrogate modeling techniques have become indispensable in accelerating the discovery and optimization of high-entropy alloys(HEAs), especially when integrating computational predictions with sparse experimental observations. This study systematically evaluates the fitting performance of four prominent surrogate models conventional Gaussian Processes(cGP), Deep Gaussian Processes(DGP), encoder-decoder neural networks for multi-output regression and XGBoost applied to a hybrid dataset of experimental and computational properties in the AlCoCrCuFeMnNiV HEA system. We specifically assess their capabilities in predicting correlated material properties, including yield strength, hardness, modulus, ultimate tensile strength, elongation, and average hardness under dynamic and quasi-static conditions, alongside auxiliary computational properties. The comparison highlights the strengths of hierarchical and deep modeling approaches in handling heteroscedastic, heterotopic, and incomplete data commonly encountered in materials informatics. Our findings illustrate that DGP infused with machine learning-based prior outperform other surrogates by effectively capturing inter-property correlations and input-dependent uncertainty. This enhanced predictive accuracy positions advanced surrogate models as powerful tools for robust and data-efficient materials design.

cs.LG↗

arXiv2025

Application-specific machine-learned interatomic potentials: exploring the trade-off between DFT convergence, MLIP expressivity, and computational cost

Machine-learned interatomic potentials (MLIPs) are revolutionizing computational materials science and chemistry by offering an efficient alternative to {\em ab initio} molecular dynamics (MD) simulations. However, fitting high-quality MLIPs remains a challenging, time-consuming, and computationally intensive task where numerous trade-offs have to be considered, e.g., How much and what kind of atomic configurations should be included in the training set? Which level of {\em ab initio} convergence should be used to generate the training set? Which loss function should be used for fitting the MLIP? Which machine learning architecture should be used to train the MLIP? The answers to these questions significantly impact both the computational cost of MLIP training and the accuracy and computational cost of subsequent MLIP MD simulations. In this study, we use a configurationally diverse beryllium dataset and quadratic spectral neighbor analysis potential. We demonstrate that joint optimization of energy versus force weights, training set selection strategies, and convergence settings of the {\em ab initio} reference simulations, as well as model complexity can lead to a significant reduction in the overall computational cost associated with training and evaluating MLIPs. This opens the door to computationally efficient generation of high-quality MLIPs for a range of applications which demand different accuracy versus training and evaluation cost trade-offs.

physics.comp-ph↗

arXiv2025

Heating of Cs2Te Photocathode via Field Emission and RF Pulsed Heating: Implication Toward Breakdown

The occurrence of radio-frequency (rf) breakdown limits operational electromagnetic gradients in accelerator structures. Experimental evidence often suggests that breakdown events are associated with temperature and dark current spikes on the surface of rf devices. In the past decade, there has been increased interest in unveiling the mechanism behind breakdown initiation in metal copper and copper alloys. However, effort regarding breakdown phenomenon in photocathode relevant semiconductors have been more limited. In this work, we explore field-emission-assisted heating via Nottingham and Joule processes, as a possible candidate for breakdown initiation. For this, field emission from intrinsic Cs2Te ultra-thin film coated on a copper substrate was modeled within the Stratton/Baskin/Lvov/Fursey (SBLF) formalism, describing the processes and effects in the bulk and on the surface of a photocathode exposed to high radio-frequency electromagnetic fields. It is shown that field emission characteristic deviates significantly from the classical Fowler-Nordheim (FN) theory, whereby predicting that dark current is orders of magnitude lower than one expected by FN law. Conventional pulsed heating was also found to impose negligible heating to the photo-cathode. Both conclusions suggest that Cs2Te photocathode coated on a metal substrate would be insensitive to catastrophic thermal-material runaway breakdown, unlike what is observed for metal surfaces. Finally, a few unconventional breakdown candidate scenarios are identified and discussed including thermo-elastic deformation and avalanche breakdown.

physics.acc-ph↗

arXiv2025

Uncertainty Quantification for Misspecified Machine Learned Interatomic Potentials

The use of high-dimensional regression techniques from machine learning has significantly improved the quantitative accuracy of interatomic potentials. Atomic simulations can now plausibly target quantitative predictions in a variety of settings, which has brought renewed interest in robust means to quantify uncertainties on simulation results. In many practical settings, encompassing both classical and a large class of machine learning potentials, the dominant form of uncertainty is currently not due to lack of training data but to misspecification, namely the inability of any one choice of model parameters to exactly match all ab initio training data. However, Bayesian inference, the most common formal tool used to quantify uncertainty, is known to ignore misspecification and thus significantly underestimates parameter uncertainties. Here, we employ a recent misspecification-aware regression technique to quantify parameter uncertainties, which is then propagated to a broad range of phase and defect properties in tungsten via brute force resampling or implicit differentiation. The propagated misspecification uncertainties robustly envelope errors to direct \textit{ab initio} calculation of material properties outside of the training dataset, an essential requirement for any quantitative multi-scale modeling scheme. Finally, we demonstrate application to recent foundational machine learning interatomic potentials, accurately predicting and bounding errors in MACE-MPA-0 energy predictions across the diverse materials project database. Perspectives for the approach in multiscale simulation workflows are discussed.

cond-mat.mtrl-sci↗

arXiv2024

Breaking the mold: overcoming the time constraints of molecular dynamics on general-purpose hardware

The evolution of molecular dynamics (MD) simulations has been intimately linked to that of computing hardware. For decades following the creation of MD, simulations have improved with computing power along the three principal dimensions of accuracy, atom count (spatial scale), and duration (temporal scale). Since the mid-2000s, computer platforms have however failed to provide strong scaling for MD as scale-out CPU and GPU platforms that provide substantial increases to spatial scale do not lead to proportional increases in temporal scale. Important scientific problems therefore remained inaccessible to direct simulation, prompting the development of increasingly sophisticated algorithms that present significant complexity, accuracy, and efficiency challenges. While bespoke MD-only hardware solutions have provided a path to longer timescales for specific physical systems, their impact on the broader community has been mitigated by their limited adaptability to new methods and potentials. In this work, we show that a novel computing architecture, the Cerebras Wafer Scale Engine, completely alters the scaling path by delivering unprecedentedly high simulation rates up to 1.144M steps/second for 200,000 atoms whose interactions are described by an Embedded Atom Method potential. This enables direct simulations of the evolution of materials using general-purpose programmable hardware over millisecond timescales, dramatically increasing the space of direct MD simulations that can be carried out.

cs.DC↗

arXiv2024

Hierarchical Gaussian Process-Based Bayesian Optimization for Materials Discovery in High Entropy Alloy Spaces

Bayesian optimization (BO) is a powerful and data-efficient method for iterative materials discovery and design, particularly valuable when prior knowledge is limited, underlying functional relationships are complex or unknown, and the cost of querying the materials space is significant. Traditional BO methodologies typically utilize conventional Gaussian Processes (cGPs) to model the relationships between material inputs and properties, as well as correlations within the input space. However, cGP-BO approaches often fall short in multi-objective optimization scenarios, where they are unable to fully exploit correlations between distinct material properties. Leveraging these correlations can significantly enhance the discovery process, as information about one property can inform and improve predictions about others. This study addresses this limitation by employing advanced kernel structures to capture and model multi-dimensional property correlations through multi-task (MTGPs) or deep Gaussian Processes (DGPs), thus accelerating the discovery process. We demonstrate the effectiveness of MTGP-BO and DGP-BO in rapidly and robustly solving complex materials design challenges that occur within the context of complex multi-objective optimization -- carried out by leveraging the pyiron workflow manager over FCC FeCrNiCoCu high entropy alloy (HEA) spaces, where traditional cGP-BO approaches fail. Furthermore, we highlight how the differential costs associated with querying various material properties can be strategically leveraged to make the materials discovery process more cost-efficient.

cond-mat.mtrl-sci↗

arXiv2024

Large radiation back-flux from Monte Carlo simulations of fusion neutron-material interactions

Radiation back-fluxes, generated from neutron-material interactions in fusion power reactors, can dramatically impact the plasma dynamics, e.g., by seeding runaway electrons during disruptions via Compton scattering of background electrons by wall-emitted gamma radiation. Here, we quantify these back-fluxes, including neutrons, gamma rays, and electrons, using Monte Carlo calculations for a range of structural material candidates and first wall thicknesses. The radiation back-flux magnitudes are remarkably large, with neutron and gamma radiation back-fluxes on the same order of magnitude as the incident fusion neutron flux. Electron back-fluxes are two orders of magnitudes lower, but are emitted at sufficiently high energies to provide a relatively large back-current through the sheath which may cause sheath reversal. Material configuration plays a key role in determining back-flux magnitudes. The structural material chiefly determines the neutron back-flux magnitude, while the first wall thickness principally attenuates the gamma ray and electron back-fluxes. In addition to prompt back-fluxes, which are emitted immediately after fusion neutrons impact the surface, significant delayed gamma ray and electron back-fluxes arise from nuclear decay processes in the activated materials. These delayed back-flux magnitudes range from 2%--7% of the prompt back-fluxes, and remain present during transients when fusion no longer occurs. During disruptions, build-up of delayed gamma radiation back-flux represents potential runaway electron seeding mechanisms, posing additional challenges for disruption mitigation in a power reactor compared with non-nuclear plasma operations. This work highlights the impact of these radiation back-fluxes plasma performance and demonstrates the importance of considering back-flux generation in materials selection for fusion power reactors.

physics.plasm-ph↗

arXiv2024

Best of Both Worlds: Enforcing Detailed Balance in Machine Learning Models of Transition Rates

The slow microstructural evolution of materials often plays a key role in determining material properties. When the unit steps of the evolution process are slow, direct simulation approaches such as molecular dynamics become prohibitive and Kinetic Monte-Carlo (kMC) algorithms, where the state-to-state evolution of the system is represented in terms of a continuous-time Markov chain, are instead frequently relied upon to efficiently predict long-time evolution. The accuracy of kMC simulations however relies on the complete and accurate knowledge of reaction pathways and corresponding kinetics. This requirement becomes extremely stringent in complex systems such as concentrated alloys where the astronomical number of local atomic configurations makes the a priori tabulation of all possible transitions impractical. Machine learning models of transition kinetics have been used to mitigate this problem by enabling the efficient on-the-fly prediction of kinetic parameters. In this study, we show how physics-informed ML architectures can exactly enforce the detailed balance condition, by construction. Using the diffusion of a vacancy in a concentrated alloy as an example, we show that such ML architectures also exhibit superior performance in terms of prediction accuracy, demonstrating that the imposition of physical constraints can facilitate the accurate learning of barriers at no increase in computational cost.

cond-mat.mtrl-sci↗

arXiv2024

Information-entropy-driven generation of material-agnostic datasets for machine-learning interatomic potentials

In contrast to their empirical counterparts, machine-learning interatomic potentials (MLIAPs) promise to deliver near-quantum accuracy over broad regions of configuration space. However, due to their generic functional forms and extreme flexibility, they can catastrophically fail to capture the properties of novel, out-of-sample configurations, making the quality of the training set a determining factor, especially when investigating materials under extreme conditions. We propose a novel automated dataset generation method based on the maximization of the information entropy of the feature distribution, aiming at an extremely broad coverage of the configuration space in a way that is agnostic to the properties of specific target materials. The ability of the dataset to capture unique material properties is demonstrated on a range of unary materials, including elements with the fcc (Al), bcc (W), hcp (Be, Re and Os), graphite (C), and trigonal (Sb, Te) ground states. MLIAPs trained to this dataset are shown to be accurate over a range of application-relevant metrics, as well as extremely robust over very broad swaths of configurations space, even without dataset fine-tuning or hyper-parameter optimization, making the approach extremely attractive to rapidly and autonomously develop general-purpose MLIAPs suitable for simulations in extreme conditions.

cond-mat.mtrl-sci↗

arXiv2024

Changes in Dislocation Punching Behavior Due to Hydrogen-Seeded Helium Bubble Growth in Tungsten

The accumulation of gas atoms in tungsten is a topic of long-standing interest to the plasma-facing materials community due the metal's use as a divertor material in some tokamak fusion reactors. The nucleation and growth of He/H gas bubbles (along with their isotopes) can result from impinging fluxes of these gases which give rise to damage at the W divertor surface. The inclusion of He or H in W has been studied extensively by the community, finding that He bubbles modify the surface through periodic dislocation punching and bursting mechanisms while H bubbles impact the metal through plastic-strain induced material failure. However, the mechanisms which are present during the combined flux of both He and H is not well-studied atomistically. Motivated by this, an atomistic modeling study is conducted using molecular dynamics to assess the behavior of mixed concentration He:H bubbles in W. We find that the introduction of H into a growing He bubble results in a dramatic change in the nature and presence of dislocation loops which are typically generated via dislocation punching in over-pressurized He bubbles. Most notably, at high H concentrations, there is a switchover in energetic favorability from glissile 1/2$<$111$>$ dislocations to sessile $<$100$>$ dislocations. This thermodynamic crossover could imply a significant reduction in W surface morphology changes than with pure He bubbles and, additionally, we show this to have implications on the trapping of H in the bubbles and their associated dislocations.

cond-mat.mtrl-sci↗