Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “equivariance”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 127 records · Page 7

Spin topological field theory and fermionic matrix product states

Here, we study state-sum constructions of G-equivariant spin topological quantum field theory (TQFTs) and their relationship to matrix product states. In the Neveu-Schwarz, Ramond, and twisted sectors, states of the TQFT are generalized matrix product states. Our findings are applied to the classification of fermionic short-range-entangled phases with a unitary symmetry G to determine the group law on the set of such phases. Interesting subtleties appear when the total symmetry group is a nontrivial extension of G by fermion parity.

36 MATERIALS SCIENCE↗

Fermionic matrix product states and one-dimensional short-range entangled phases with antiunitary symmetries

We extend the formalism of matrix product states (MPS) to describe one-dimensional gapped systems of fermions with both unitary and antiunitary symmetries. Additionally, systems with orientation-reversing spatial symmetries are considered. The short-ranged entangled phases of such systems are classified by three invariants, which characterize the projective action of the symmetry on edge states. We give interpretations of these invariants as properties of states on the closed chain. The relationship between fermionic MPS systems at a renormalization group fixed point and equivariant algebras is exploited to derive a group law for the stacking of fermionic phases. Furthermore, the result generalizes known classifications to symmetry groups that are nontrivial extensions of fermion parity and time-reversal.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Flow matching beyond kinematics: Generating jets with particle identification and trajectory displacement information

We introduce the first generative model trained on the etlass dataset. Our model generates jets at the constituent level, and it is a permutation-equivariant continuous normalizing flow (CNF) trained with the flow matching technique. It is conditioned on the jet type, so that a single model can be used to generate the ten different jet types of etlass. For the first time, we also introduce a generative model that goes beyond the kinematic features of jet constituents. The etlass dataset includes more features, such as particle-ID and track impact parameter, and we demonstrate that our CNF can accurately model all of these additional features as well. Our generative model for etlass expands on the versatility of existing jet generation techniques, enhancing their potential utility in high-energy physics research, and offering a more comprehensive understanding of the generated jets. Published by the American Physical Society 2025

Birk, Joschka (ORCID:0000000219310127)↗

Learning broken symmetries with approximate invariance

Recognizing symmetries in data allows for significant boosts in neural network training, which is especially important where training data are limited. In many cases, however, the exact underlying symmetry is present only in an idealized dataset, and is broken in actual data, due to asymmetries in the detector, or varying response resolution as a function of particle momentum. Standard approaches, such as data augmentation or equivariant networks fail to represent the nature of the full, broken symmetry, effectively overconstraining the response of the neural network. We propose a learning model which balances the generality and asymptotic performance of unconstrained networks with the rapid learning of constrained networks. This is achieved through a dual-subnet structure, where one network is constrained by the symmetry and the other is not, along with a learned symmetry factor. In a simplified toy example that demonstrates violation of Lorentz invariance, our model learns as rapidly as symmetry constrained networks but escapes its performance limitations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Noncommutative vortices and flux tubes from Yang-Mills theories with spontaneously generated fuzzy extra dimensions

We consider a U(2) Yang-Mills theory on MxS{sub F}{sup 2}, where M is an arbitrary noncommutative manifold, and S{sub F}{sup 2} is a fuzzy sphere spontaneously generated from a noncommutative U(N) Yang-Mills theory on M, coupled to a triplet of scalars in the adjoint of U(N). Employing the SU(2)-equivariant gauge field constructed in [D. Harland and S. Kurkcuoglu, Nucl. Phys. B 821, 380 (2009).], we perform the dimensional reduction of the theory over the fuzzy sphere. The emergent model is a noncommutative U(1) gauge theory coupled adjointly to a set of scalar fields. We study this model on the Groenewald-Moyal plane M=R{sub {theta}}{sup 2} and find that, in certain limits, it admits noncommutative, non-Bogomol'nyi-Prasad-Somerfield vortex as well as flux-tube (fluxon) solutions and discuss some of their properties.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Chern-Simons flows on Aloff-Wallach spaces and spin(7) instantons

Because of their explicit construction, Aloff-Wallach spaces are prominent in flux compactifications. They carry G{sub 2} structures and admit the G{sub 2}-instanton equations, which are natural Bogomol'nyi-Prasad-Sommerfeld equations for Yang-Mills instantons on seven-manifolds and extremize a Chern-Simons-type functional. We consider the Chern-Simons flow between different G{sub 2} instantons on Aloff-Wallach spaces, which is equivalent to spin(7) instantons on a cylinder over them. For a general SU(3)-equivariant gauge connection, the generalized instanton equations turn into gradient-flow equations on C{sup 3}xR{sup 2}, with a particular cubic superpotential. For the simplest member of the Aloff-Wallach family (with 3-Sasakian structure) we present an explicit instanton solution of tanh-like shape.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Liquid-liquid phase transition of hydrogen and its critical point: Analysis from ab initio simulation and a machine-learned potential

We simulate high-pressure hydrogen in its liquid phase close to molecular dissociation using a machine-learned interatomic potential. The model is trained with density functional theory (DFT) forces and energies, with the Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional. We show that an accurate NequIP model, an E(3)-equivariant neural network potential, accurately reproduces the phase transition present in PBE. Moreover, the computational efficiency of this model allows for substantially longer molecular dynamics trajectories, enabling us to perform a finite-size scaling (FSS) analysis to distinguish between a crossover and a true first-order phase transition. Here, we locate the critical point of this transition, the liquid-liquid phase transition (LLPT), at 1200-1300 K and 155-160 GPa, a temperature lower than most previous estimates and close to the melting transition.

08 HYDROGEN↗

Point-particle drag, lift, and torque closure models using machine learning: Hierarchical approach and interpretability

Developing deterministic neighborhood-informed point-particle closure models using machine learning has garnered interest recently from the dispersed multiphase flow community. The robustness of neural models for this complex multibody problem is hindered by the availability of particle-resolved data. Here, the present work addresses this unavoidable limitation of data paucity by implementing two strategies: (1) by using a rotation and reflection equivariant neural network and (2) by pursuing a physics-based hierarchical machine learning approach. The resulting machine-learned models are observed to achieve a maximum accuracy of 85% and 96% in the prediction of neighbor-induced force and torque fluctuations, respectively, for a wide range of Reynolds number and volume fraction conditions considered. Furthermore, we pursue force and torque network architectures that provide universal prediction spanning a wide range of Reynolds number (0.25 ≤ Re ≤250) and particle volume fraction (0 ≤ φ ≤0.4). The hierarchical nature of the approach enables improved prediction of quantities such as streamwise torque, by going beyond binary interactions to include trinary interactions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Central Charges of Liouville and Toda Theories from M5-Branes

We show that the central charge of the Liouville and Toda theories of type A, D, and E can be reproduced by equivariantly integrating the anomaly eight-form of the corresponding six-dimensional N=(0,2) theories, which describe the low-energy dynamics of M5-branes.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Wilson loops with neural networks

Wilson loops are essential objects in QCD and have been pivotal in scale setting and demonstrating confinement. Various generalizations are crucial for computations needed in effective field theories. In lattice gauge theory, Wilson loop calculations face challenges, including excited-state contamination at short times and the signal-to-noise ratio issue at longer times. To address these problems, we develop a new method by using neural networks to parametrize interpolators for the static quark-antiquark pair. We construct gauge-equivariant layers for the network and train it to find the ground state of the system. The trained network itself is then treated as our new observable for the inference. Our results demonstrate a significant improvement in the signal compared to traditional Wilson loops, performing as well as Coulomb-gauge Wilson-line correlators while maintaining gauge invariance. Additionally, we present an example where the optimized ground state is used to measure the static force directly, as well as another example combining this method with the multilevel algorithm. Finally, we extend the formalism to find excited-state interpolators for static quark-antiquark systems. To our knowledge, this work is the first study of neural networks with a physically motivated loss function for Wilson loops.

Bellscheidt, Verena [Massachusetts Inst. of Techno↗

Geometric deep learning of RNA structure

RNA molecules adopt three-dimensional structures that are critical to their function and of interest in drug discovery. Few RNA structures are known, however, and predicting them computationally has proven challenging. We introduce a machine learning approach that enables identification of accurate structural models without assumptions about their defining characteristics, despite being trained with only 18 known RNA structures. The resulting scoring function, the Atomic Rotationally Equivariant Scorer (ARES), substantially outperforms previous methods and consistently produces the best results in community-wide blind RNA structure prediction challenges. By learning effectively even from a small amount of data, our approach overcomes a major limitation of standard deep neural networks. Because it uses only atomic coordinates as inputs and incorporates no RNA-specific information, this approach is applicable to diverse problems in structural biology, chemistry, materials science, and beyond.

Townshend, Raphael J. L.↗

Instantons and Chern-Simons flows in 6, 7 and 8 dimensions

The existence of K-instantons on a cylinder M{sup 7} = ℝ{sub τ} × K/H over a homogeneous nearly Käller 6-manifold K/H requires a conformally parallel or a cocalibrated G{sub 2}-structure on M{sup 7}. The generalized anti-self-duality on M{sup 7} implies a Chern-Simons flow on K/H which runs between instantons on the coset. For K-equivariant connections, the torsionful Yang-Mills equation reduces to a particular quartic dynamics for a Newtonian particle on ℂ. When the torsion corresponds to one of the G{sub 2}-structures, this dynamics follows from a gradient or hamiltonian flow equation, respectively. We present the analytic (kink-type) solutions and plot numerical non-BPS solutions for general torsion values interpolating between the instantonic ones.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Pure Yang–Mills Solutions on dS{sub 4}

We consider pure SU(2) Yang–Mills theory on four-dimensional de Sitter space dS{sub 4} and construct smooth and spatially homogeneous classical Yang–Mills fields. Slicing dS{sub 4} as R×S{sup 3}, via an SU(2)-equivariant ansatz we reduce the Yang–Mills equations to ordinary matrix differential equations and further to Newtonian dynamics in a particular three-dimensional potential. Its classical trajectories yield spatially homogeneous Yang–Mills solutions in a very simple explicit form, depending only on de Sitter time with an exponential decay in the past and future. These configurations have not only finite energy, but their action is also finite and bounded from below. We present explicit coordinate representations of the simplest examples (for the fundamental SU(2) representation). Instantons (Yang–Mills solutions on the Wick-rotated S{sup 4} ) and solutions on AdS{sub 4} are also briefly discussed.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Euclidean neural networks (e3nn) v1.0

E(3) is the Euclidean group in dimension 3. That is the group of rotations, translations and mirror. e3nn is a pytorch library that aims to create E(3) equivariant neural networks.

Lapchevskyi, Kostiantyn↗

e3nn tutorial v0.2

Repository and website for tutorials on 3d Euclidean equivariant neural networks.

Smidt, Tess↗

HydraGNN v3.0

New or improved capabilities included in v3.0 release are as follows: 1. Enhancement in message passing layers through generalization of the class inheritance to enable the inclusion of a broader set of message passing policies Inclusion of equivariant message passing layers from the original implementations of: SchNet (https://pubs.aip.org/aip/jcp/article/148/24/241722/962591/SchNet-A-deep-learning-architecture-for-molecules); DimeNet++ (https://arxiv.org/abs/2011.14115); EGNN models (https://arxiv.org/pdf/2102.09844.pdf) 2. Restructuring of class inheritance for data management 3. Support of DDStore https://github.com/ORNL/DDStore capabilities for improved distributed data parallelism on large volumes of data that cannot fit on intra-node memory capacities 4. Large-scale system support for OLCF-Crusher and OLCF-Frontier

Lupo Pasini, Massimiliano [Oak Ridge National Labo↗

SIDDA: SInkhorn Dynamic Domain Adaptation for Image Classification

SInkhorn Dynamic Domain Adaptation (SIDDA) supplements the experiments presented in 2501.14048, SIDDA: SInkhorn Dynamic Domain Adaptation for Image Classification with Equivariant Neural Networks. SIDDA introduces a semi-supervised, automatic domain adaptation method that leverages Sinkhorn divergences to dynamically adjust the regularization in the optimal transport plan and the weighting between classification and domain adaptation loss terms during training.

Pandya, Sneh [Fermi National Accelerator Laborator↗

HydraGNN v4.0

The new version of HydraGNN v4.0 provides additional core capabilities, such as: Inclusion of multi-body atomistic cluster expansion MACE, polarizable atom interaction neural network PAINN, and equivariant principal neighborhood aggregation (PNAEq) among the message passing layers supported -Inclusion of graph transformers to directly model long-range interactions between nodes that are distant in the graph topology Integration of graph transformers with message passing layers by combining the graph embedding generated by the two mechanisms, which allows for an improved expressivity of the HydraGNN architecture Improved re-implementation of multi-task learning (MTL) to allow its use for stabilized training across imbalanced, multi-source, multi-fidelity data Introduction of multi-task parallelism, a newly proposed type of model parallelism specifically for MTL architectures, which allows to dispatch different output decoding heads to different GPU devices Integration of multi-task parallelism with pre-existing distributed data parallelism to enable a 2D parallelization for distributed training Improved portability of the distributed training across Intel GPUs, which has been testes on ALCF exascale supercomputer Aurora Inclusion of 2-level fine-grained energy profilers portable across NVIDIA, AMD, and Intel GPUs to monitor the power and energy consumption associated with different functions executed by the HydraGNN code during data pre-load and training Restructuring of previous examples and inclusion of new sets of examples to illustrate the download, preprocess, and training of HydraGNN models on new large-scale open-source datasets for atomistic materials modeling (e.g., Alexandria, Transition1x, OMat24, OMol25)

Lupo Pasini, Massimiliano [Oak Ridge National Labo↗