Search

Recent studies have demonstrated the strong empirical performance of diffusion models on discrete sequences across domains from natural language to biological sequence generation. For example, in the protein inverse folding task, conditional diffusion models have achieved impressive results in generating natural-like sequences that fold back into the original structure. However, practical design tasks often require not only modeling a conditional distribution but also optimizing specific task objectives. For instance, we may prefer protein sequences with high stability. To address this, we consider the scenario where we have pre-trained discrete diffusion models that can generate natural-like sequences, as well as reward models that map sequences to task objectives. We then formulate the reward maximization problem within discrete diffusion models, analogous to reinforcement learning (RL), while minimizing the KL divergence against pretrained diffusion models to preserve naturalness. To solve this RL problem, we propose a novel algorithm, DRAKES, that enables direct backpropagation of rewards through entire trajectories generated by diffusion models, by making the originally non-differentiable trajectories differentiable using the Gumbel-Softmax trick. Our theoretical analysis indicates that our approach can generate sequences that are both natural-like and yield high rewards. While similar tasks have been recently explored in diffusion models for continuous domains, our work addresses unique algorithmic and theoretical challenges specific to discrete diffusion models, which arise from their foundation in continuous-time Markov chains rather than Brownian motion. Finally, we demonstrate the effectiveness of DRAKES in generating DNA and protein sequences that optimize enhancer activity and protein stability, respectively, important tasks for gene therapies and protein-based therapeutics.

Graph representation learning, involving both node features and graph structures, is crucial for real-world applications but often encounters pervasive noise. State-of-the-art methods typically address noise by focusing separately on node features with large language models (LLMs) and on graph structures with graph structure learning models (GSLMs). In this paper, we introduce LangGSL, a robust framework that integrates the complementary strengths of pre-trained language models and GSLMs to jointly enhance both node feature and graph structure learning. In LangGSL, we first leverage LLMs to filter noise in the raw data and extract valuable cleaned information as features, enhancing the synergy of downstream models. During the mutual learning phase in LangGSL, the core idea is to leverage the relatively small language model (LM) to process local attributes and generate reliable pseudo-labels and informative node embeddings, which are then integrated into the GSLM's prediction phase. This approach enriches the global context and enhances overall performance. Meanwhile, GSLM refines the evolving graph structure constructed from the LM's output, offering updated labels back to the LM as additional guidance, thus facilitating a more effective mutual learning process. The LM and GSLM work synergistically, complementing each other's strengths and offsetting weaknesses within a variational information-maximizing framework, resulting in enhanced node features and a more robust graph structure. Extensive experiments on diverse graph datasets of varying scales and across different task scenarios demonstrate the scalability and effectiveness of the proposed approach., Comment: Graph structure learning, Graph representation learning, Large language models, Graph neural networks

Full Waveform Inversion (FWI) is a vital technique for reconstructing high-resolution subsurface velocity maps from seismic waveform data, governed by partial differential equations (PDEs) that model wave propagation. Traditional machine learning approaches typically map seismic data to velocity maps by encoding seismic waveforms into latent embeddings and decoding them into velocity maps. In this paper, we introduce a novel framework that reframes FWI as a joint diffusion process in a shared latent space, bridging seismic waveform data and velocity maps. Our approach has two key components: first, we merge the bottlenecks of two separate autoencoders-one for seismic data and one for velocity maps-into a unified latent space using vector quantization to establish a shared codebook. Second, we train a diffusion model in this latent space, enabling the simultaneous generation of seismic and velocity map pairs by sampling and denoising the latent representations, followed by decoding each modality with its respective decoder. Remarkably, our jointly generated seismic-velocity pairs approximately satisfy the governing PDE without any additional constraint, offering a new geometric interpretation of FWI. The diffusion process learns to score the latent space according to its deviation from the PDE, with higher scores representing smaller deviations from the true solutions. By following this diffusion process, the model traces a path from random initialization to a valid solution of the governing PDE. Our experiments on the OpenFWI dataset demonstrate that the generated seismic and velocity map pairs not only exhibit high fidelity and diversity but also adhere to the physical constraints imposed by the governing PDE.

Recent technological advancements have enhanced our ability to collect and analyze rich multimodal data (e.g., speech, video, and eye gaze) to better inform learning and training experiences. While previous reviews have focused on parts of the multimodal pipeline (e.g., conceptual models and data fusion), a comprehensive literature review on the methods informing multimodal learning and training environments has not been conducted. This literature review provides an in-depth analysis of research methods in these environments, proposing a taxonomy and framework that encapsulates recent methodological advances in this field and characterizes the multimodal domain in terms of five modality groups: Natural Language, Video, Sensors, Human-Centered, and Environment Logs. We introduce a novel data fusion category -- mid fusion -- and a graph-based technique for refining literature reviews, termed citation graph pruning. Our analysis reveals that leveraging multiple modalities offers a more holistic understanding of the behaviors and outcomes of learners and trainees. Even when multimodality does not enhance predictive accuracy, it often uncovers patterns that contextualize and elucidate unimodal data, revealing subtleties that a single modality may miss. However, there remains a need for further research to bridge the divide between multimodal learning and training studies and foundational AI research., Comment: Submitted to ACM Computing Surveys. Currently under review

Charge-spin interconversion processes underpin the generation of spin-orbit torques in magnetic/nonmagnetic bilayers. However, efficient sources of spin currents such as 5d metals are also efficient spin sinks, resulting in a large increase of magnetic damping. Here we show that a partially-oxidized 3d metal can generate a strong orbital torque without a significant increase in damping. Measurements of the torque efficiency {\xi} and Gilbert damping {\alpha} in CoFe/CuOx and CoFe/Pt indicate that {\xi} is comparable. The increase in damping relative to a single CoFe layer is {\Delta}{\alpha}<0.002 in CoFe/CuOx and {\Delta}{\alpha} ~ 0.005 - 0.02 in CoFe/Pt, depending on CoFe thickness. We ascribe the nonreciprocal relationship between {\Delta}{\alpha} and {\xi} in CoFe/CuOx to the small orbital-to-spin current ratio generated by magnetic resonance in CoFe and the lack of an efficient spin sink in CuOx. Our findings provide new perspectives on the efficient excitation of magnetization dynamics via the orbital torque.

Recent global estimates suggest that as many as 2.41 billion individuals have health conditions that would benefit from rehabilitation services. Home-based Physical Therapy (PT) faces significant challenges in providing interactive feedback and meaningful observation for therapists and patients. To fill this gap, we present MicroXercise, which integrates micro-motion analysis with wearable sensors, providing therapists and patients with a comprehensive feedback interface, including video, text, and scores. Crucially, it employs multi-dimensional Dynamic Time Warping (DTW) and attribution-based explainable methods to analyze the existing deep learning neural networks in monitoring exercises, focusing on a high granularity of exercise. This synergistic approach is pivotal, providing output matching the input size to precisely highlight critical subtleties and movements in PT, thus transforming complex AI analysis into clear, actionable feedback. By highlighting these micro-motions in different metrics, such as stability and range of motion, MicroXercise significantly enhances the understanding and relevance of feedback for end-users. Comparative performance metrics underscore its effectiveness over traditional methods, such as a 39% and 42% improvement in Feature Mutual Information (FMI) and Continuity. MicroXercise is a step ahead in home-based physical therapy, providing a technologically advanced and intuitively helpful solution to enhance patient care and outcomes., Comment: Accepted by IEEE/ACM CHASE 2024

Exceptional points with coalescence of eigenvalues and eigenvectors are spectral singularities in the parameter space, achieving which often needs fine-tuning of parameters in quantum systems. We predict a \textit{persistent} realization of nodal magnon-photon polariton, i.e., a polariton of long wavelength without any gap splitting in a thin ferromagnetic insulator film sandwiched by two normal metals, which persistently exists when the ferromagnet is sufficiently thick $\sim 100$~nm due to the joint effect of dissipation and dissipative coupling. We perform the model calculation \textit{beyond the perturbation theory} using a classical approach, develop a quantum scheme able to account for the Ohmic dissipation, and find ultrastrong coupling with coupling strength comparable to the bare magnon frequency. Via revealing a simple conversion relation we extend this formalism to superconductors and predict the gap opened by the ultrastrong coupling strongly depends on the direction of polariton propagation. Our findings may help search for robust non-Hermitian topological phases in magnonic and spintronic devices., Comment: 18 pages, 7 figures

Graph Neural Networks (GNNs) are vital in data science but are increasingly susceptible to adversarial attacks. To help researchers develop more robust GNN models, it's essential to focus on designing strong attack models as foundational benchmarks and guiding references. Among adversarial attacks, gray-box poisoning attacks are noteworthy due to their effectiveness and fewer constraints. These attacks exploit GNNs' need for retraining on updated data, thereby impacting their performance by perturbing these datasets. However, current research overlooks the real-world scenario of incomplete graphs.To address this gap, we introduce the Robust Incomplete Deep Attack Framework (RIDA). It is the first algorithm for robust gray-box poisoning attacks on incomplete graphs. The approach innovatively aggregates distant vertex information and ensures powerful data utilization.Extensive tests against 9 SOTA baselines on 3 real-world datasets demonstrate RIDA's superiority in handling incompleteness and high attack performance on the incomplete graph.

Ultrasound computed tomography (USCT) is a promising technique that achieves superior medical imaging reconstruction resolution by fully leveraging waveform information, outperforming conventional ultrasound methods. Despite its advantages, high-quality USCT reconstruction relies on extensive data acquisition by a large number of transducers, leading to increased costs, computational demands, extended patient scanning times, and manufacturing complexities. To mitigate these issues, we propose a new USCT method called APS-USCT, which facilitates imaging with sparse data, substantially reducing dependence on high-cost dense data acquisition. Our APS-USCT method consists of two primary components: APS-wave and APS-FWI. The APS-wave component, an encoder-decoder system, preprocesses the waveform data, converting sparse data into dense waveforms to augment sample density prior to reconstruction. The APS-FWI component, utilizing the InversionNet, directly reconstructs the speed of sound (SOS) from the ultrasound waveform data. We further improve the model's performance by incorporating Squeeze-and-Excitation (SE) Blocks and source encoding techniques. Testing our method on a breast cancer dataset yielded promising results. It demonstrated outstanding performance with an average Structural Similarity Index (SSIM) of 0.8431. Notably, over 82% of samples achieved an SSIM above 0.8, with nearly 61% exceeding 0.85, highlighting the significant potential of our approach in improving USCT image reconstruction by efficiently utilizing sparse data., Comment: MICCAI

The magnetic resonance of iron garnets is at the heart of rich physics and provides a crucial technology used inside microwave components. A barrier still stands in the way of integrated cryogenic devices with iron garnets, because their epitaxy requires paramagnetic substrates that cause large microwave losses at low temperatures. Finding alternative combinations of substrates and magnetic garnets is thus essential to enabling cryogenic integrated magnonics. With this motivation, we develop a new approach to the synthesis of epitaxial garnet films, based on the co-sputtering of binary oxides, enabling a precise and continuous tuning of the iron garnet composition. By substituting a controlled proportion of the iron with additional yttrium, we obtain Y$_{3}$(Y$_{x}$Fe$_{5-x}$)O$_{12}$ films of high crystalline quality that can be lattice-matched to yttrium scandium gallium garnet (YSGG), a diamagnetic substrate suitable for low-temperature applications. Ferromagnetic resonance performed at low temperatures confirms the elimination of substrate paramagnetism losses on YSGG. We also show that the Y$_{3}$(Y$_{x}$Fe$_{5-x}$)O$_{12}$ system can be matched to other substrates, such as gadolinium yttrium scandium gallium garnet and gadolinium scandium gallium garnet. Additionally, we demonstrate that Bi-substituted films of (Bi$_{0.8}$Y$_{2.2}$)Fe$_{5}$O$_{12}$ also offer ideal lattice matching to YSGG for low-temperature magnetic resonance. The present approach of controlled substitutions in iron garnets to obtain ideal lattice matching with diamagnetic substrates opens a very promising pathway to magnonic devices operating in low-temperature environments., Comment: Main: 18 pages, 6 figures; Supplement: 9 pages, 6 figures

Spatial-Temporal Graph (STG) data is characterized as dynamic, heterogenous, and non-stationary, leading to the continuous challenge of spatial-temporal graph learning. In the past few years, various GNN-based methods have been proposed to solely focus on mimicking the relationships among node individuals of the STG network, ignoring the significance of modeling the intrinsic features that exist in STG system over time. In contrast, modern Selective State Space Models (SSSMs) present a new approach which treat STG Network as a system, and meticulously explore the STG system's dynamic state evolution across temporal dimension. In this work, we introduce Spatial-Temporal Graph Mamba (STG-Mamba) as the first exploration of leveraging the powerful selective state space models for STG learning by treating STG Network as a system, and employing the Spatial-Temporal Selective State Space Module (ST-S3M) to precisely focus on the selected STG latent features. Furthermore, to strengthen GNN's ability of modeling STG data under the setting of selective state space models, we propose Kalman Filtering Graph Neural Networks (KFGN) for dynamically integrate and upgrade the STG embeddings from different temporal granularities through a learnable Kalman Filtering statistical theory-based approach. Extensive empirical studies are conducted on three benchmark STG forecasting datasets, demonstrating the performance superiority and computational efficiency of STG-Mamba. It not only surpasses existing state-of-the-art methods in terms of STG forecasting performance, but also effectively alleviate the computational bottleneck of large-scale graph networks in reducing the computational cost of FLOPs and test inference time. The implementation code is available at: \url{https://github.com/LincanLi98/STG-Mamba}.

Deep generative models (DGMs) have been widely developed for graph data. However, much less investigation has been carried out on understanding the latent space of such pretrained graph DGMs. These understandings possess the potential to provide constructive guidelines for crucial tasks, such as graph controllable generation. Thus in this work, we are interested in studying this problem and propose GraphCG, a method for the unsupervised discovery of steerable factors in the latent space of pretrained graph DGMs. We first examine the representation space of three pretrained graph DGMs with six disentanglement metrics, and we observe that the pretrained representation space is entangled. Motivated by this observation, GraphCG learns the steerable factors via maximizing the mutual information between semantic-rich directions, where the controlled graph moving along the same direction will share the same steerable factors. We quantitatively verify that GraphCG outperforms four competitive baselines on two graph DGMs pretrained on two molecule datasets. Additionally, we qualitatively illustrate seven steerable factors learned by GraphCG on five pretrained DGMs over five graph datasets, including two for molecules and three for point clouds.

The Hanle magnetoresistance is a telltale signature of spin precession in nonmagnetic conductors, in which strong spin-orbit coupling generates edge spin accumulation via the spin Hall effect. Here, we report the existence of a large Hanle magnetoresistance in single layers of Mn with weak spin-orbit coupling, which we attribute to the orbital Hall effect. The simultaneous observation of a sizable Hanle magnetoresistance and vanishing small spin Hall magnetoresistance in BiYIG/Mn bilayers corroborates the orbital origin of both effects. We estimate an orbital Hall angle of 0.016, an orbital relaxation time of 2 ps and diffusion length of the order of 2 nm in disordered Mn. Our findings indicate that current-induced orbital moments are responsible for magnetoresistance effects comparable to or even larger than those determined by spin moments, and provide a tool to investigate nonequilibrium orbital transport phenomena.

Full-waveform inversion (FWI) plays a vital role in geoscience to explore the subsurface. It utilizes the seismic wave to image the subsurface velocity map. As the machine learning (ML) technique evolves, the data-driven approaches using ML for FWI tasks have emerged, offering enhanced accuracy and reduced computational cost compared to traditional physics-based methods. However, a common challenge in geoscience, the unprivileged data, severely limits ML effectiveness. The issue becomes even worse during model pruning, a step essential in geoscience due to environmental complexities. To tackle this, we introduce the EdGeo toolkit, which employs a diffusion-based model guided by physics principles to generate high-fidelity velocity maps. The toolkit uses the acoustic wave equation to generate corresponding seismic waveform data, facilitating the fine-tuning of pruned ML models. Our results demonstrate significant improvements in SSIM scores and reduction in both MAE and MSE across various pruning ratios. Notably, the ML model fine-tuned using data generated by EdGeo yields superior quality of velocity maps, especially in representing unprivileged features, outperforming other existing methods.

We report the experimental observation of large spin pumping signals in YIG/Pt system driven by broad-wavevector spin-wave spin current. 280 nm-wide microwave inductive antennas offer broad-wavevector excitation which, in combination with quasi-flatband of YIG, allows a large number of magnons to participate in spin pumping at a given frequency. Through comparison with ferromagnetic resonance spin pumping, we attribute the enhancement of the spin current to the multichromatic magnons. The high efficiency of spin current generation enables us to uncover nontrivial propagating properties in ultra-low power regions. Additionally, our study achieves the spatially separated detection of magnons, allowing the direct extraction of the decay length. The synergistic combination of the capability of broad-wavevector excitation, enhanced voltage signals, and nonlocal detection provides a new avenue for the electrical exploration of spin waves dynamics.

The carbon capture, utilization, and storage (CCUS) framework is an essential component in reducing greenhouse gas emissions, with its success hinging on the comprehensive knowledge of subsurface geology and geomechanics. Passive seismic event relocation and fault detection serve as indispensable tools, offering vital insights into subsurface structures and fluid migration pathways. Accurate identification and localization of seismic events, however, face significant challenges, including the necessity for high-quality seismic data and advanced computational methods. To address these challenges, we introduce a novel deep learning method, DeFault, specifically designed for passive seismic source relocation and fault delineating for passive seismic monitoring projects. By leveraging data domain-adaptation, DeFault allows us to train a neural network with labeled synthetic data and apply it directly to field data. Using DeFault, the passive seismic sources are automatically clustered based on their recording time and spatial locations, and subsequently, faults and fractures are delineated accordingly. We demonstrate the efficacy of DeFault on a field case study involving CO2 injection related microseismic data from the Decatur, Illinois area. Our approach accurately and efficiently relocated passive seismic events, identified faults and aided in the prevention of potential geological hazards. Our results highlight the potential of DeFault as a valuable tool for passive seismic monitoring, emphasizing its role in ensuring CCUS project safety. This research bolsters the understanding of subsurface characterization in CCUS, illustrating machine learning's capacity to refine these methods. Ultimately, our work bear significant implications for CCUS technology deployment, an essential strategy in combating climate change.

In micro-seismic event measurements, pinpointing the passive source's exact spatial and temporal location is paramount. This research advocates for the combined use of both P- and S-wave data, captured by geophone monitoring systems, to improve source inversion accuracy. Drawing inspiration from the secondary source concept in Elastic Reflection Full Waveform Inversion (ERFWI), we introduce an equivalent source term. This term combines source functions and source images. Our optimization strategy iteratively refines the spatial locations of the source, its temporal functions, and associated velocities using a full waveform inversion framework. Under the premise of an isotropic medium with consistent density, the source is defined by two spatial and three temporal components. This offers a nuanced source representation in contrast to the conventional seismic moment tensor. To address gradient computation, we employ the adjoint-state method. However, we encountered pronounced non-linearity in waveform inversion of micro-seismic events, primarily due to the unknown source origin time, resulting in cycle skipping challenges. To counteract this, we devised an objective function that is decoupled from the source origin time. This function is formulated by convolving reference traces with both observed and predicted data. Through the concurrent inversion of the source image, source time function, and velocity model, our method offers precise estimations of these parameters, as validated by a synthetic 2D example based on a modified Marmousi model. This nested inversion approach promises enhanced accuracy in determining the source image, time function, and velocity model.

Steganography, or hiding messages in plain sight, is a form of information hiding that is most commonly used for covert communication. As modern steganographic mediums include images, text, audio, and video, this communication method is being increasingly used by bad actors to propagate malware, exfiltrate data, and discreetly communicate. Current protection mechanisms rely upon steganalysis, or the detection of steganography, but these approaches are dependent upon prior knowledge, such as steganographic signatures from publicly available tools and statistical knowledge about known hiding methods. These dependencies render steganalysis useless against new or unique hiding methods, which are becoming increasingly common with the application of deep learning models. To mitigate the shortcomings of steganalysis, this work focuses on a deep learning sanitization technique called SUDS that is not reliant upon knowledge of steganographic hiding techniques and is able to sanitize universal and dependent steganography. SUDS is tested using least significant bit method (LSB), dependent deep hiding (DDH), and universal deep hiding (UDH). We demonstrate the capabilities and limitations of SUDS by answering five research questions, including baseline comparisons and an ablation study. Additionally, we apply SUDS to a real-world scenario, where it is able to increase the resistance of a poisoned classifier against attacks by 1375%., Comment: Accepted to European Conference on Artificial Intelligence (ECAI) 2023

Ultrasound computed tomography (USCT) is an emerging imaging modality that holds great promise for breast imaging. Full-waveform inversion (FWI)-based image reconstruction methods incorporate accurate wave physics to produce high spatial resolution quantitative images of speed of sound or other acoustic properties of the breast tissues from USCT measurement data. However, the high computational cost of FWI reconstruction represents a significant burden for its widespread application in a clinical setting. The research reported here investigates the use of a convolutional neural network (CNN) to learn a mapping from USCT waveform data to speed of sound estimates. The CNN was trained using a supervised approach with a task-informed loss function aiming at preserving features of the image that are relevant to the detection of lesions. A large set of anatomically and physiologically realistic numerical breast phantoms (NBPs) and corresponding simulated USCT measurements was employed during training. Once trained, the CNN can perform real-time FWI image reconstruction from USCT waveform data. The performance of the proposed method was assessed and compared against FWI using a hold-out sample of 41 NBPs and corresponding USCT data. Accuracy was measured using relative mean square error (RMSE), structural self-similarity index measure (SSIM), and lesion detection performance (DICE score). This numerical experiment demonstrates that a supervised learning model can achieve accuracy comparable to FWI in terms of RMSE and SSIM, and better performance in terms of task performance, while significantly reducing computational time., Comment: 13 pages, 12 figures

This paper investigates the impact of big data on deep learning models to help solve the full waveform inversion (FWI) problem. While it is well known that big data can boost the performance of deep learning models in many tasks, its effectiveness has not been validated for FWI. To address this gap, we present an empirical study that investigates how deep learning models in FWI behave when trained on OpenFWI, a collection of large-scale, multi-structural, synthetic datasets published recently. In particular, we train and evaluate the FWI models on a combination of 10 2D subsets in OpenFWI that contain 470K pairs of seismic data and velocity maps in total. Our experiments demonstrate that training on the combined dataset yields an average improvement of 13.03% in MAE, 7.19% in MSE and 1.87% in SSIM compared to each split dataset, and an average improvement of 28.60%, 21.55% and 8.22% in the leave-one-out generalization test. We further demonstrate that model capacity needs to scale in accordance with data size for optimal improvement, where our largest model yields an average improvement of 20.06%, 13.39% and 0.72% compared to the smallest one.

Elastic geophysical properties (such as P- and S-wave velocities) are of great importance to various subsurface applications like CO$_2$ sequestration and energy exploration (e.g., hydrogen and geothermal). Elastic full waveform inversion (FWI) is widely applied for characterizing reservoir properties. In this paper, we introduce $\mathbf{\mathbb{E}^{FWI}}$, a comprehensive benchmark dataset that is specifically designed for elastic FWI. $\mathbf{\mathbb{E}^{FWI}}$ encompasses 8 distinct datasets that cover diverse subsurface geologic structures (flat, curve, faults, etc). The benchmark results produced by three different deep learning methods are provided. In contrast to our previously presented dataset (pressure recordings) for acoustic FWI (referred to as OpenFWI), the seismic dataset in $\mathbf{\mathbb{E}^{FWI}}$ has both vertical and horizontal components. Moreover, the velocity maps in $\mathbf{\mathbb{E}^{FWI}}$ incorporate both P- and S-wave velocities. While the multicomponent data and the added S-wave velocity make the data more realistic, more challenges are introduced regarding the convergence and computational cost of the inversion. We conduct comprehensive numerical experiments to explore the relationship between P-wave and S-wave velocities in seismic data. The relation between P- and S-wave velocities provides crucial insights into the subsurface properties such as lithology, porosity, fluid content, etc. We anticipate that $\mathbf{\mathbb{E}^{FWI}}$ will facilitate future research on multiparameter inversions and stimulate endeavors in several critical research topics of carbon-zero and new energy exploration. All datasets, codes and relevant information can be accessed through our website at https://efwi-lanl.github.io/, Comment: 20 pages, 11 figures

We report the experimental observation of the spin-wave moir\'e edge and cavity modes using Brillouin light scattering spectro-microscopy in a nanostructured magnetic moir\'e lattice consisting of two twisted triangle antidot lattices based on an yttrium iron garnet thin film. Spin-wave moir\'e edge modes are detected at an optimal twist angle and with a selective excitation frequency. At a given twist angle, the magnetic field acts as an additional degree of freedom for tuning the chiral behavior of the magnon edge modes. Micromagnetic simulations indicate that the edge modes emerge within the original magnonic band gap and at the intersection between a mini-flatband and a propagation magnon branch. Our theoretical estimate for the Berry curvature of the magnon-magnon coupling suggests a non-trivial topology for the chiral edge modes and confirms the key role played by the dipolar interaction. Our findings shed light on the topological nature of the magnon edge mode for emergent moir\'e magnonics.

We report on coherent propagation of antiferromagnetic (AFM) spin waves over a long distance ($\sim$10 $\mu$m) at room temperature in a canted AFM $\alpha$-Fe$_2$O$_3$ with the Dzyaloshinskii-Moriya interaction (DMI). Unprecedented high group velocities (up to 22.5 km/s) are characterized by microwave transmission using all-electrical spin wave spectroscopy. We derive analytically AFM spin-wave dispersion in the presence of the DMI which accounts for our experimental results. The AFM spin waves excited by nanometric coplanar waveguides with large wavevectors enter the exchange regime and follow a quasi-linear dispersion relation. Fitting of experimental data with our theoretical model yields an AFM exchange stiffness length of 1.7 angstrom. Our results provide key insights on AFM spin dynamics and demonstrate high-speed functionality for AFM magnonics.

Physical therapy (PT) is crucial for patients to restore and maintain mobility, function, and well-being. Many on-site activities and body exercises are performed under the supervision of therapists or clinicians. However, the postures of some exercises at home cannot be performed accurately due to the lack of supervision, quality assessment, and self-correction. Therefore, in this paper, we design a new framework, PhysiQ, that continuously tracks and quantitatively measures people's off-site exercise activity through passive sensory detection. In the framework, we create a novel multi-task spatio-temporal Siamese Neural Network that measures the absolute quality through classification and relative quality based on an individual's PT progress through similarity comparison. PhysiQ digitizes and evaluates exercises in three different metrics: range of motions, stability, and repetition., Comment: 22 pages

In the study of subsurface seismic imaging, solving the acoustic wave equation is a pivotal component in existing models. The advancement of deep learning enables solving partial differential equations, including wave equations, by applying neural networks to identify the mapping between the inputs and the solution. This approach can be faster than traditional numerical methods when numerous instances are to be solved. Previous works that concentrate on solving the wave equation by neural networks consider either a single velocity model or multiple simple velocity models, which is restricted in practice. Instead, inspired by the idea of operator learning, this work leverages the Fourier neural operator (FNO) to effectively learn the frequency domain seismic wavefields under the context of variable velocity models. We also propose a new framework paralleled Fourier neural operator (PFNO) for efficiently training the FNO-based solver given multiple source locations and frequencies. Numerical experiments demonstrate the high accuracy of both FNO and PFNO with complicated velocity models in the OpenFWI datasets. Furthermore, the cross-dataset generalization test verifies that PFNO adapts to out-of-distribution velocity models. Moreover, PFNO has robust performance in the presence of random noise in the labels. Finally, PFNO admits higher computational efficiency on large-scale testing datasets than the traditional finite-difference method. The aforementioned advantages endow the FNO-based solver with the potential to build powerful models for research on seismic waves., Comment: 26 pages, 17 figures

A leading nonlinear effect in magnonics is the interaction that splits a high-frequency magnon into two low-frequency ones with conserved linear momentum. Here, we report experimental observation of nonlocal three-magnon scattering between spatially separated magnetic systems, viz. a CoFeB nanowire and an yttrium iron garnet (YIG) thin film. Above a certain threshold power of an applied microwave field, a CoFeB Kittel magnon splits into a pair of counter-propagating YIG magnons that induce voltage signals in Pt electrodes on each side, in excellent agreement with model calculations based on the interlayer dipolar interaction. The excited YIG magnon pairs reside mainly in the first excited (n=1) perpdendicular standing spin-wave mode. With increasing power, the n=1 magnons successively scatter into nodeless (n=0) magnons through a four-magnon process. Our results help to assess non-local scattering processes in magnonic circuits that may enable quantum entanglement between distant magnons for quantum information applications.

Graph Self-Supervised Learning (GSSL) provides a robust pathway for acquiring embeddings without expert labelling, a capability that carries profound implications for molecular graphs due to the staggering number of potential molecules and the high cost of obtaining labels. However, GSSL methods are designed not for optimisation within a specific domain but rather for transferability across a variety of downstream tasks. This broad applicability complicates their evaluation. Addressing this challenge, we present "Molecular Graph Representation Evaluation" (MOLGRAPHEVAL), generating detailed profiles of molecular graph embeddings with interpretable and diversified attributes. MOLGRAPHEVAL offers a suite of probing tasks grouped into three categories: (i) generic graph, (ii) molecular substructure, and (iii) embedding space properties. By leveraging MOLGRAPHEVAL to benchmark existing GSSL methods against both current downstream datasets and our suite of tasks, we uncover significant inconsistencies between inferences drawn solely from existing datasets and those derived from more nuanced probing. These findings suggest that current evaluation methodologies fail to capture the entirety of the landscape., Comment: Camera ready, NeurIPS Benchmark 2023

Graph Neural Networks (GNNs) are effective tools for graph representation learning. Most GNNs rely on a recursive neighborhood aggregation scheme, named message passing, thereby their theoretical expressive power is limited to the first-order Weisfeiler-Lehman test (1-WL). An effective approach to this challenge is to explicitly retrieve some annotated examples used to enhance GNN models. While retrieval-enhanced models have been proved to be effective in many language and vision domains, it remains an open question how effective retrieval-enhanced GNNs are when applied to graph datasets. Motivated by this, we want to explore how the retrieval idea can help augment the useful information learned in the graph neural networks, and we design a retrieval-enhanced scheme called GRAPHRETRIEVAL, which is agnostic to the choice of graph neural network models. In GRAPHRETRIEVAL, for each input graph, similar graphs together with their ground-true labels are retrieved from an existing database. Thus they can act as a potential enhancement to complete various graph property predictive tasks. We conduct comprehensive experiments over 13 datasets, and we observe that GRAPHRETRIEVAL is able to reach substantial improvements over existing GNNs. Moreover, our empirical study also illustrates that retrieval enhancement is a promising remedy for alleviating the long-tailed label distribution problem., Comment: Accepted by ECAI 2023

Subgraph matching is a fundamental problem in various fields that use graph structured data. Subgraph matching algorithms enumerate all isomorphic embeddings of a query graph q in a data graph G. An important branch of matching algorithms exploit the backtracking search approach which recursively extends intermediate results following a matching order of query vertices. It has been shown that the matching order plays a critical role in time efficiency of these backtracking based subgraph matching algorithms. In recent years, many advanced techniques for query vertex ordering (i.e., matching order generation) have been proposed to reduce the unpromising intermediate results according to the preset heuristic rules. In this paper, for the first time we apply the Reinforcement Learning (RL) and Graph Neural Networks (GNNs) techniques to generate the high-quality matching order for subgraph matching algorithms. Instead of using the fixed heuristics to generate the matching order, our model could capture and make full use of the graph information, and thus determine the query vertex order with the adaptive learning-based rule that could significantly reduces the number of redundant enumerations. With the help of the reinforcement learning framework, our model is able to consider the long-term benefits rather than only consider the local information at current ordering step.Extensive experiments on six real-life data graphs demonstrate that our proposed matching order generation technique could reduce up to two orders of magnitude of query processing time compared to the state-of-the-art algorithms., Comment: Accepted by ICDE 2022

Artificial intelligence (AI) provides a promising substitution for streamlining COVID-19 diagnoses. However, concerns surrounding security and trustworthiness impede the collection of large-scale representative medical data, posing a considerable challenge for training a well-generalised model in clinical practices. To address this, we launch the Unified CT-COVID AI Diagnostic Initiative (UCADI), where the AI model can be distributedly trained and independently executed at each host institution under a federated learning framework (FL) without data sharing. Here we show that our FL model outperformed all the local models by a large yield (test sensitivity /specificity in China: 0.973/0.951, in the UK: 0.730/0.942), achieving comparable performance with a panel of professional radiologists. We further evaluated the model on the hold-out (collected from another two hospitals leaving out the FL) and heterogeneous (acquired with contrast materials) data, provided visual explanations for decisions made by the model, and analysed the trade-offs between the model performance and the communication costs in the federated training process. Our study is based on 9,573 chest computed tomography scans (CTs) from 3,336 patients collected from 23 hospitals located in China and the UK. Collectively, our work advanced the prospects of utilising federated learning for privacy-preserving AI in digital health., Comment: Nature Machine Intelligence

Full waveform inversion (FWI) is widely used in geophysics to reconstruct high-resolution velocity maps from seismic data. The recent success of data-driven FWI methods results in a rapidly increasing demand for open datasets to serve the geophysics community. We present OpenFWI, a collection of large-scale multi-structural benchmark datasets, to facilitate diversified, rigorous, and reproducible research on FWI. In particular, OpenFWI consists of 12 datasets (2.1TB in total) synthesized from multiple sources. It encompasses diverse domains in geophysics (interface, fault, CO2 reservoir, etc.), covers different geological subsurface structures (flat, curve, etc.), and contains various amounts of data samples (2K - 67K). It also includes a dataset for 3D FWI. Moreover, we use OpenFWI to perform benchmarking over four deep learning methods, covering both supervised and unsupervised learning regimes. Along with the benchmarks, we implement additional experiments, including physics-driven methods, complexity analysis, generalization study, uncertainty quantification, and so on, to sharpen our understanding of datasets and methods. The studies either provide valuable insights into the datasets and the performance, or uncover their current limitations. We hope OpenFWI supports prospective research on FWI and inspires future open-source efforts on AI for science. All datasets and related information can be accessed through our website at https://openfwi-lanl.github.io/, Comment: This manuscript has been accepted by NeurIPS 2022 dataset and benchmark track

In this paper, we consider the problem of iterative machine teaching, where a teacher provides examples sequentially based on the current iterative learner. In contrast to previous methods that have to scan over the entire pool and select teaching examples from it in each iteration, we propose a label synthesis teaching framework where the teacher randomly selects input teaching examples (e.g., images) and then synthesizes suitable outputs (e.g., labels) for them. We show that this framework can avoid costly example selection while still provably achieving exponential teachability. We propose multiple novel teaching algorithms in this framework. Finally, we empirically demonstrate the value of our framework., Comment: NeurIPS 2021 Spotlight (v5: 28 pages, 20 figures, fixed typos in v4)

Molecular graph representation learning is a fundamental problem in modern drug and material discovery. Molecular graphs are typically modeled by their 2D topological structures, but it has been recently discovered that 3D geometric information plays a more vital role in predicting molecular functionalities. However, the lack of 3D information in real-world scenarios has significantly impeded the learning of geometric graph representation. To cope with this challenge, we propose the Graph Multi-View Pre-training (GraphMVP) framework where self-supervised learning (SSL) is performed by leveraging the correspondence and consistency between 2D topological structures and 3D geometric views. GraphMVP effectively learns a 2D molecular graph encoder that is enhanced by richer and more discriminative 3D geometry. We further provide theoretical insights to justify the effectiveness of GraphMVP. Finally, comprehensive experiments show that GraphMVP can consistently outperform existing graph SSL methods.

Among the biggest challenges we face in utilizing neural networks trained on waveform data (i.e., seismic, electromagnetic, or ultrasound) is its application to real data. The requirement for accurate labels forces us to develop solutions using synthetic data, where labels are readily available. However, synthetic data often do not capture the reality of the field/real experiment, and we end up with poor performance of the trained neural network (NN) at the inference stage. We describe a novel approach to enhance supervised training on synthetic data with real data features (domain adaptation). Specifically, for tasks in which the absolute values of the vertical axis (time or depth) of the input data are not crucial, like classification, or can be corrected afterward, like velocity model building using a well-log, we suggest a series of linear operations on the input so the training and application data have similar distributions. This is accomplished by applying two operations on the input data to the NN model: 1) The crosscorrelation of the input data (i.e., shot gather, seismic image, etc.) with a fixed reference trace from the same dataset. 2) The convolution of the resulting data with the mean (or a random sample) of the autocorrelated data from another domain. In the training stage, the input data are from the synthetic domain and the auto-correlated data are from the real domain, and random samples from real data are drawn at every training epoch. In the inference/application stage, the input data are from the real subset domain and the mean of the autocorrelated sections are from the synthetic data subset domain. Example applications on passive seismic data for microseismic event source location determination and active seismic data for predicting low frequencies are used to demonstrate the power of this approach in improving the applicability of trained models to real data., Comment: 27 pages, 15 figures