Search

Basecalling, an essential step in many genome analysis studies, relies on large Deep Neural Networks (DNNs) to achieve high accuracy. Unfortunately, these DNNs are computationally slow and inefficient, leading to considerable delays and resource constraints in the sequence analysis process. A Computation-In-Memory (CIM) architecture using memristors can significantly accelerate the performance of DNNs. However, inherent device non-idealities and architectural limitations of such designs can greatly degrade the basecalling accuracy, which is critical for accurate genome analysis. To facilitate the adoption of memristor-based CIM designs for basecalling, it is important to (1) conduct a comprehensive analysis of potential CIM architectures and (2) develop effective strategies for mitigating the possible adverse effects of inherent device non-idealities and architectural limitations. This paper proposes Swordfish, a novel hardware/software co-design framework that can effectively address the two aforementioned issues. Swordfish incorporates seven circuit and device restrictions or non-idealities from characterized real memristor-based chips. Swordfish leverages various hardware/software co-design solutions to mitigate the basecalling accuracy loss due to such non-idealities. To demonstrate the effectiveness of Swordfish, we take Bonito, the state-of-the-art (i.e., accurate and fast), open-source basecaller as a case study. Our experimental results using Sword-fish show that a CIM architecture can realistically accelerate Bonito for a wide range of real datasets by an average of 25.7x, with an accuracy loss of 6.01%., Comment: To appear in 56th IEEE/ACM International Symposium on Microarchitecture (MICRO), 2023

Recent research and development in the mid-infrared (IR) wavelength range (2-20 um) for a variety of applications, such as trace gas monitoring, thermal imaging, and free space communications have shown tremendous and fascinating progress. MXenes, which mainly refer to two-dimensional (2D) transition-metal carbides, nitrides, and carbonitrides, have drawn a lot of interest since their first investigation in 2011. MXenes project enormous potential for use in optoelectronics, photonics, catalysis, and energy harvesting fields proven by extensive experimental and theoretical studies over a decade. MXenes offers a novel 2D nano platform for cutting-edge optoelectronics devices due to their interesting mechanical, optical, and electrical capabilities, along with their elemental and chemical composition. We here discuss the key developments of MXene emphasizing the evolution of material synthesis methods over time and the resulting device applications. Photonic and optoelectronic device design and fabrication for mid-IR photonics are demonstrated by integrating MXene materials with various electrical and photonic platforms. Here, we show the potential of using Mxene in photonics for mid-IR applications and a pathway toward achieving next-generation devices for various applications., Comment: 50 Pages, 21 figures

Neural video codecs have recently become competitive with standard codecs such as HEVC in the low-delay setting. However, most neural codecs are large floating-point networks that use pixel-dense warping operations for temporal modeling, making them too computationally expensive for deployment on mobile devices. Recent work has demonstrated that running a neural decoder in real time on mobile is feasible, but shows this only for 720p RGB video. This work presents the first neural video codec that decodes 1080p YUV420 video in real time on a mobile device. Our codec relies on two major contributions. First, we design an efficient codec that uses a block-based motion compensation algorithm available on the warping core of the mobile accelerator, and we show how to quantize this model to integer precision. Second, we implement a fast decoder pipeline that concurrently runs neural network components on the neural signal processor, parallel entropy coding on the mobile GPU, and warping on the warping core. Our codec outperforms the previous on-device codec by a large margin with up to 48% BD-rate savings, while reducing the MAC count on the receiver side by $10 \times$. We perform a careful ablation to demonstrate the effect of the introduced motion compensation scheme, and ablate the effect of model quantization., Comment: Matches version published at WACV 2024

In this work, we show that, for any simply-connected elliptic space $S$ admitting a pure minimal Sullivan model with a differential of constant length, we have ${\rm TC}_0(S)\leq 2{\rm cat}_0(S)+\chi_{\pi}(S)$ where $\chi_{\pi}(S)$ is the homotopy characteristic. This is a consequence of a structure theorem for this type of models, which is actually our main result.

This paper aims to explore the long-term behavior of some nonlocal high-order-in-time wave equations. These equations, which have come to be known as Moore--Gibson--Thompson equations, arise in the context of acoustic wave propagation when taking into account thermal relaxation mechanisms in complex media such as human tissue. While the long-term behavior of linear local-in-time acoustic equations is well understood, their nonlocal counterparts still retain many mysteries. We establish here a set of assumptions that ensures exponential decay of the energy of the system. These assumptions are then shown to be verified by a large class of rapidly decaying memory kernels. Under weaker assumptions on the kernel we show that one may still obtain that the energy vanishes but without a rate of convergence. Furthermore, we refine previous results on the local well-posedness of the studied equation and establish a necessary initial-data compatibility condition for the solvability of the problem.

Research in the field of Materials Science and Engineering focuses on the design, synthesis, properties, and performance of materials. An important class of materials that is widely investigated are crystalline materials, including metals and semiconductors. Crystalline material typically contains a distinct type of defect called "dislocation". This defect significantly affects various material properties, including strength, fracture toughness, and ductility. Researchers have devoted a significant effort in recent years to understanding dislocation behavior through experimental characterization techniques and simulations, e.g., dislocation dynamics simulations. This paper presents how data from dislocation dynamics simulations can be modeled using semantic web technologies through annotating data with ontologies. We extend the already existing Dislocation Ontology by adding missing concepts and aligning it with two other domain-related ontologies (i.e., the Elementary Multi-perspective Material Ontology and the Materials Design Ontology) allowing for representing the dislocation simulation data efficiently. Moreover, we show a real-world use case by representing the discrete dislocation dynamics data as a knowledge graph (DisLocKG) that illustrates the relationship between them. We also developed a SPARQL endpoint that brings extensive flexibility to query DisLocKG.

Image camouflage has been utilized to create clean-label poisoned images for implanting backdoor into a DL model. But there exists a crucial limitation that one attack/poisoned image can only fit a single input size of the DL model, which greatly increases its attack budget when attacking multiple commonly adopted input sizes of DL models. This work proposes to constructively craft an attack image through camouflaging but can fit multiple DL models' input sizes simultaneously, namely OmClic. Thus, through OmClic, we are able to always implant a backdoor regardless of which common input size is chosen by the user to train the DL model given the same attack budget (i.e., a fraction of the poisoning rate). With our camouflaging algorithm formulated as a multi-objective optimization, M=5 input sizes can be concurrently targeted with one attack image, which artifact is retained to be almost visually imperceptible at the same time. Extensive evaluations validate the proposed OmClic can reliably succeed in various settings using diverse types of images. Further experiments on OmClic based backdoor insertion to DL models show that high backdoor performances (i.e., attack success rate and clean data accuracy) are achievable no matter which common input size is randomly chosen by the user to train the model. So that the OmClic based backdoor attack budget is reduced by M$\times$ compared to the state-of-the-art camouflage based backdoor attack as a baseline. Significantly, the same set of OmClic based poisonous attack images is transferable to different model architectures for backdoor implant.

The open problems related to cosmological tensions in current times have opened new paths to study new probes to constrain cosmological parameters in standard and extended cosmologies, in particular, to determine at a local level the value of the Hubble constant $H_0$, through independent techniques. However, while standard Cosmological Constant Cold Dark Matter ($\Lambda$CDM) model has been well constrained and parts of extended cosmology have been intensively studied, the physics behind them aspects restrains our possibilities of selecting the best cosmological model that can show a significant difference from the first model. Therefore, to explore a possible deviation from a such model that can explain the current discrepancy on the $H_0$ value, in this work we consider adding the current local observables, e.g. Supernovae Type Ia (SNIa), $H(z)$ measurements, and Baryon Acoustic Observations (BAO) combined with two new calibrated Quasars (QSO) datasets using ultraviolet, x-ray and optical plane techniques. While these can be identified as part of the high-redshift standard candle objects, the main characteristics of these are based on fluxes distributions calibrated up to $z \sim 7 $. We consider five $H_0$ prior scenarios to develop these calibrations. Furthermore, we found that our estimations provide the possibility to relax the $H_0$ tension at 2$\sigma$ using a QSO ultraviolet sample in combination with late measurements showing higher values of $H_0$. Our results can be an initial start for more serious treatments in the quasars physics from ultraviolet, x-ray, and optical plane techniques behind the local observations as cosmological probes to relax the cosmological tensions problems., Comment: 29 pages, 11 figures and 16 table. Version accepted in Physics of the Dark Universe

Physical evolution of cosmological models can be tested by using expansion data, while growth history of these models is capable of testing dynamics of the inhomogeneous parts of energy density. The growth factor, as well as its growth index, gives a clear indication of the performance of cosmological models in the regime of structure formation of early Universe. In this work, we explore the growth index in several leading $f(T)$ cosmological models, based on a specific class of teleparallel gravity theories. These have become prominent in the literature and lead to other formulations of teleparallel gravity. Here we adopt a generalized approach by obtaining the M\'{e}sz\'{a}ros equation without immediately imposing the subhorizon limit, because this assumption could lead to over-simplification. This approach gives avenue to study at which $k$ modes the subhorizon limit starts to apply. We obtain numerical results for the growth factor and growth index for a variety of data set combinations for each $f(T)$ model.

Face parsing infers a pixel-wise label map for each semantic facial component. Previous methods generally work well for uncovered faces, however, they overlook facial occlusion and ignore some contextual areas outside a single face, especially when facial occlusion has become a common situation during the COVID-19 epidemic. Inspired by the lighting phenomena in everyday life, where illumination from four distinct lamps provides a more uniform distribution than a single central light source, we propose a novel homogeneous tanh-transform for image preprocessing, which is made up of four tanh-transforms. These transforms fuse the central vision and the peripheral vision together. Our proposed method addresses the dilemma of face parsing under occlusion and compresses more information from the surrounding context. Based on homogeneous tanh-transforms, we propose an occlusion-aware convolutional neural network for occluded face parsing. It combines information in both Tanh-polar space and Tanh-Cartesian space, capable of enhancing receptive fields. Furthermore, we introduce an occlusion-aware loss to focus on the boundaries of occluded regions. The network is simple, flexible, and can be trained end-to-end. To facilitate future research of occluded face parsing, we also contribute a new cleaned face parsing dataset. This dataset is manually purified from several academic or industrial datasets, including CelebAMask-HQ, Short-video Face Parsing, and the Helen dataset, and will be made public. Experiments demonstrate that our method surpasses state-of-the-art methods in face parsing under occlusion.

We study the phenomenology of the charged Higgs boson at future muon colliders. We investigate both the pair production $\mu^+ \mu^- \to H^+ H^-$, the single production $\mu^+ \mu^- \to W^\pm H^\mp$, as well as the Vector Boson Fusion (VBF) $\{e^+e^-, \mu^+ \mu^-\} \to \nu \bar{\nu} H^+ H^-$. We show that the neutral Higgs exchange diagrams in the muon collider case can lead to a significant boost in the cross sections through their Yukawa couplings. Our results for the muon collider are systematically compared to the corresponding ones at $e^+e^-$ machines. It is demonstrated that the vector boson fusion (VBF) $e^+e^- \to \nu \bar{\nu} H^+ H^-$ can compete with the mentioned $2\to 2$ processes. We select benchmark points and perform signal-background analyses, taking into account detector simulations. We demonstrate the discovery region at $5\sigma$ and the excluded region at $2\sigma$ levels at a 3 TeV muon collider., Comment: 43 pages, 14 figures, 10 tables, accepted to be published in PRD

Graph unlearning emerges as a crucial advancement in the pursuit of responsible AI, providing the means to remove sensitive data traces from trained models, thereby upholding the right to be forgotten. It is evident that graph machine learning exhibits sensitivity to data privacy and adversarial attacks, necessitating the application of graph unlearning techniques to address these concerns effectively. In this comprehensive survey paper, we present the first systematic review of graph unlearning approaches, encompassing a diverse array of methodologies and offering a detailed taxonomy and up-to-date literature overview to facilitate the understanding of researchers new to this field. Additionally, we establish the vital connections between graph unlearning and differential privacy, augmenting our understanding of the relevance of privacy-preserving techniques in this context. To ensure clarity, we provide lucid explanations of the fundamental concepts and evaluation measures used in graph unlearning, catering to a broader audience with varying levels of expertise. Delving into potential applications, we explore the versatility of graph unlearning across various domains, including but not limited to social networks, adversarial settings, and resource-constrained environments like the Internet of Things (IoT), illustrating its potential impact in safeguarding data privacy and enhancing AI systems' robustness. Finally, we shed light on promising research directions, encouraging further progress and innovation within the domain of graph unlearning. By laying a solid foundation and fostering continued progress, this survey seeks to inspire researchers to further advance the field of graph unlearning, thereby instilling confidence in the ethical growth of AI systems and reinforcing the responsible application of machine learning techniques in various domains., Comment: 22 page review paper on graph unlearning

Recurrent neural networks (RNNs) are known to be universal approximators of dynamic systems under fairly mild and general assumptions. However, RNNs usually suffer from the issues of vanishing and exploding gradients in standard RNN training. Reservoir computing (RC), a special RNN where the recurrent weights are randomized and left untrained, has been introduced to overcome these issues and has demonstrated superior empirical performance especially in scenarios where training samples are extremely limited. On the other hand, the theoretical grounding to support this observed performance has yet been fully developed. In this work, we show that RC can universally approximate a general linear time-invariant (LTI) system. Specifically, we present a clear signal processing interpretation of RC and utilize this understanding in the problem of approximating a generic LTI system. Under this setup, we analytically characterize the optimum probability density function for configuring (instead of training and/or randomly generating) the recurrent weights of the underlying RNN of the RC. Extensive numerical evaluations are provided to validate the optimality of the derived distribution for configuring the recurrent weights of the RC to approximate a general LTI system. Our work results in clear signal processing-based model interpretability of RC and provides theoretical explanation/justification for the power of randomness in randomly generating instead of training RC's recurrent weights. Furthermore, it provides a complete optimum analytical characterization for configuring the untrained recurrent weights, marking an important step towards explainable machine learning (XML) to incorporate domain knowledge for efficient learning., Comment: This work has been accepted to IEEE Journal of Selected Topics in Signal Processing (JSTSP)

In order to characterize in detail the charge density wave (CDW) transition of 1$T$-VSe$_2$, its electronic structure and lattice dynamics are comprehensively studied by means of x-ray diffraction, angle resolved photoemission (ARPES), diffuse and inelastic x-ray scattering (IXS), and state-of-the-art first principles density functional theory calculations. Resonant elastic x-ray scattering (REXS) does not show any resonant enhancement at either V or Se K-edges, indicating that the CDW peak describes a purely structural modulation of the electronic ordering. ARPES identifies (i) a pseudogap at T$>$T$_{CDW}$, which leads to a depletion of the density of states in the $ML-M'L'$ plane at T$<$T$_{CDW}$, and (ii) anomalies in the electronic dispersion reflecting a sizable impact of phonons on it. A diffuse scattering precursor, characteristic of soft phonons, is observed at room temperature (RT) and leads to the full collapse of the low-energy phonon ($\omega_1$) with propagation vector (0.25 0 -0.3) r.l.u. We show that the frequency and linewidth of this mode are anisotropic in momentum space, reflecting the momentum dependence of the electron-phonon interaction (EPI), hence demonstrating that the origin of the CDW is, to a much larger extent, due to the momentum dependence EPI with a small contribution from nesting. The pressure dependence of the $\omega_1$ soft mode remains nearly constant up to 13 GPa at RT, with only a modest softening before the transition to the high-pressure monoclinic $C2/m$ phase. The wide set of experimental data are well captured by our state-of-the art first-principles anharmonic calculations with the inclusion of van der Waals (vdW) corrections in the exchange-correlation functional. The description of the electronics and dynamics of VSe$_2$ reported here adds important pieces of information to the understanding of the electronic modulations of TMDs.

Using high resolution tender-x-ray resonant inelastic scattering and hard-x-ray non-resonant inelastic scattering beyond the dipole limit we were able to detect electronic excitations in intermetallic UGa$_2$ that are highly atomic in nature. Analysis of the spectral lineshape reveals that the local $5f^2$ configuration characterizes the correlated nature of this ferromagnet. The orientation and directional dependence of the spectra indicate that the ground state is made of the $\Gamma_1$ singlet and/or $\Gamma_6$ doublet symmetry. With the ordered moment in the $ab$ plane, we infer that the magnetism originates from the higher lying $\Gamma_6$ doublet being mixed with the $\Gamma_1$ singlet due to inter-site exchange, qualifying UGa$_2$ to be a true quantum magnet. The ability to observe atomic excitations is crucial to resolve the on-going debate about the degree of localization versus itineracy in U intermetallics., Comment: 9 pages, 7 figures

This is an overview on the {source operator method} which leads to the construction of symmetry breaking differential operators (SBDO) in the context of tensor product of two principals series representations for the conformal group of a simple real Jordan algebra. This method can be applied to other geometric contexts: in the construction of SBDO for differential forms and for spinors, and also for the construction of Juhl's operators corresponding to the restriction from the sphere $S^n$ to $S^{n-1}$.

Users and businesses are increasingly deploying Internet of Things (IoT) devices at home, at work, and in factories. At the same time, we see an increase in the use of IPv6 for Internet connectivity. Even though the IoT ecosystem has been the focus of recent studies, there is no comprehensive analysis of IoT end-hosts in the IPv6 Internet to date. In this paper we perform an in-depth analysis of IPv6-reachable IoT hosts using active measurements. We run measurements targeting 530M IPv6 addresses on six popular IoT-related protocols. With 36.4K hosts in 156 countries we find 380x fewer IoT-speaking end-hosts compared to IPv4. Moreover, we conduct a security analysis for TLS-enabled IoT-speaking hosts identifying up to 57% untrusted certificates, with up to 32% being self-signed and 25% being expired. Finally, we plan to publish our measurement results, tools, and a website dashboard to foster further research in the area.

Extracorporeal membrane oxygenation (ECMO) is an essential life-supporting modality for COVID-19 patients who are refractory to conventional therapies. However, the proper treatment decision has been the subject of significant debate and it remains controversial about who benefits from this scarcely available and technically complex treatment option. To support clinical decisions, it is a critical need to predict the treatment need and the potential treatment and no-treatment responses. Targeting this clinical challenge, we propose Treatment Variational AutoEncoder (TVAE), a novel approach for individualized treatment analysis. TVAE is specifically designed to address the modeling challenges like ECMO with strong treatment selection bias and scarce treatment cases. TVAE conceptualizes the treatment decision as a multi-scale problem. We model a patient's potential treatment assignment and the factual and counterfactual outcomes as part of their intrinsic characteristics that can be represented by a deep latent variable model. The factual and counterfactual prediction errors are alleviated via a reconstruction regularization scheme together with semi-supervision, and the selection bias and the scarcity of treatment cases are mitigated by the disentangled and distribution-matched latent space and the label-balancing generative strategy. We evaluate TVAE on two real-world COVID-19 datasets: an international dataset collected from 1651 hospitals across 63 countries, and a institutional dataset collected from 15 hospitals. The results show that TVAE outperforms state-of-the-art treatment effect models in predicting both the propensity scores and factual outcomes on heterogeneous COVID-19 datasets. Additional experiments also show TVAE outperforms the best existing models in individual treatment effect estimation on the synthesized IHDP benchmark dataset.

The ability to characterise the three-dimensional microstructure of multiphase materials is essential for understanding the interaction between phases and associated materials properties. Here, laboratory-based diffraction-contrast tomography (DCT), a recently-established materials characterization technique that can determine grain phases, morphologies, positions and orientations in a voxel-based reconstruction method, was used to map part of a dual-phase steel alloy sample. To assess the resulting microstructures that were produced by the DCT technique, an EBSD map was collected within the same sample volume. To identify the 2D slice of the 3D DCT reconstruction that best corresponded to the EBSD map, a novel registration technique based solely on grain-averaged orientations was developed -- this registration technique requires very little a priori knowledge of dataset alignment and can be extended to other techniques that only recover grain-averaged orientation data such as far-field 3D X-ray diffraction microscopy. Once the corresponding 2D slice was identified in the DCT dataset, comparisons of phase balance, grain size, shape and texture were performed between DCT and EBSD techniques. More complicated aspects of the microstructural morphology such as grain boundary shape and grains less than a critical size were poorly reproduced by the DCT reconstruction, primarily due to the difference in resolutions of the technique compared with EBSD. However, lab-based DCT is shown to accurately determine the centre-of-mass position, orientation, and size of the large grains for each phase present, austenite and martensitic ferrite. The results reveals a complex ferrite grain network of similar crystal orientations that are absent from the EBSD dataset. Such detail demonstrates that lab-based DCT, as a technique, shows great promise in the field of multi-phase material characterization., Comment: 15 pages, 11 figures. Preprint submitted to Materials Characterization

In this work we study the cosmological perturbations in Bahamonde-Dialektopoulos-Levi Said (BDLS) theory, i.e. the teleparallel analog of Horndeski gravity. In order to understand the evolution of structure in a cosmological model, it is necessary to study its cosmology not only in the background but also perturbatively. Both Horndeski and its teleparallel analog have been analyzed a lot in the literature, but in order to study them quantitatively, we need to know their cosmological perturbations. That is why, we study here the scalar-vector-tensor decomposition of the theory and we also express the so-called alpha parameters in terms of the arbitrary functions of the theory, that designate the deviation from the {\Lambda}CDM model. We have explored tensor, vector and scalar perturbation of the action up to second order, which drastically opens up new possibilities on searches in the parameter space of scalar-tensor theories in the context of observations., Comment: 20 pages

The integral of a function $f$ defined on a symmetric space $M \simeq G/K$ may be expressed in the form of a determinant (or Pfaffian), when $f$ is $K$-invariant and, in a certain sense, a tensor power of a positive function of a single variable. The paper presents a few examples of this idea and discusses future extensions. Specifically, the examples involve symmetric cones, Grassmann manifolds, and classical domains.

The Dark Energy Spectroscopic Instrument (DESI) completed its five-month Survey Validation in May 2021. Spectra of stellar and extragalactic targets from Survey Validation constitute the first major data sample from the DESI survey. This paper describes the public release of those spectra, the catalogs of derived properties, and the intermediate data products. In total, the public release includes good-quality spectral information from 466,447 objects targeted as part of the Milky Way Survey, 428,758 as part of the Bright Galaxy Survey, 227,318 as part of the Luminous Red Galaxy sample, 437,664 as part of the Emission Line Galaxy sample, and 76,079 as part of the Quasar sample. In addition, the release includes spectral information from 137,148 objects that expand the scope beyond the primary samples as part of a series of secondary programs. Here, we describe the spectral data, data quality, data products, Large-Scale Structure science catalogs, access to the data, and references that provide relevant background to using these spectra., Comment: 43 pages, 7 figures, 17 tables, accepted for publication in the Astronomical Journal

The Dark Energy Spectroscopic Instrument (DESI) was designed to conduct a survey covering 14,000 deg$^2$ over five years to constrain the cosmic expansion history through precise measurements of Baryon Acoustic Oscillations (BAO). The scientific program for DESI was evaluated during a five month Survey Validation (SV) campaign before beginning full operations. This program produced deep spectra of tens of thousands of objects from each of the stellar (MWS), bright galaxy (BGS), luminous red galaxy (LRG), emission line galaxy (ELG), and quasar target classes. These SV spectra were used to optimize redshift distributions, characterize exposure times, determine calibration procedures, and assess observational overheads for the five-year program. In this paper, we present the final target selection algorithms, redshift distributions, and projected cosmology constraints resulting from those studies. We also present a `One-Percent survey' conducted at the conclusion of Survey Validation covering 140 deg$^2$ using the final target selection algorithms with exposures of a depth typical of the main survey. The Survey Validation indicates that DESI will be able to complete the full 14,000 deg$^2$ program with spectroscopically-confirmed targets from the MWS, BGS, LRG, ELG, and quasar programs with total sample sizes of 7.2, 13.8, 7.46, 15.7, and 2.87 million, respectively. These samples will allow exploration of the Milky Way halo, clustering on all scales, and BAO measurements with a statistical precision of 0.28% over the redshift interval $z<1.1$, 0.39% over the redshift interval $1.1

Machine learning provides a valuable tool for analyzing high-dimensional functional neuroimaging data, and is proving effective in predicting various neurological conditions, psychiatric disorders, and cognitive patterns. In functional magnetic resonance imaging (MRI) research, interactions between brain regions are commonly modeled using graph-based representations. The potency of graph machine learning methods has been established across myriad domains, marking a transformative step in data interpretation and predictive modeling. Yet, despite their promise, the transposition of these techniques to the neuroimaging domain has been challenging due to the expansive number of potential preprocessing pipelines and the large parameter search space for graph-based dataset construction. In this paper, we introduce NeuroGraph, a collection of graph-based neuroimaging datasets, and demonstrated its utility for predicting multiple categories of behavioral and cognitive traits. We delve deeply into the dataset generation search space by crafting 35 datasets that encompass static and dynamic brain connectivity, running in excess of 15 baseline methods for benchmarking. Additionally, we provide generic frameworks for learning on both static and dynamic graphs. Our extensive experiments lead to several key observations. Notably, using correlation vectors as node features, incorporating larger number of regions of interest, and employing sparser graphs lead to improved performance. To foster further advancements in graph-based data driven neuroimaging analysis, we offer a comprehensive open-source Python package that includes the benchmark datasets, baseline implementations, model training, and standard evaluation., Comment: NeurIPS23