Memory Complexity Research Articles

A Mesh-based Simulation Framework using Automatic Code Generation

Optimized parallel implementations on GPU or CPU have dramatically enhanced the fidelity, resolution and accuracy of physical simulations and mesh-based algorithms. However, attaining optimal performance requires expert knowledge and might demand complex code and memory layout optimizations. This adds to the fact that physical simulation algorithms require the implementation of derivatives, which can be a tedious and error-prone process. In recent years, researchers and practitioners have investigated the concept of designing systems that allow for a more expressive definition of mesh-based simulation code. These systems leverage domain-specific languages (DSL), automatic differentiation or symbolic computing to enhance readability of implementations without compromising performance. We follow this line of work and propose a symbolic code generation approach tailored to mesh-based computations on parallel devices. Our system extends related work by incorporating collision handling and a data access synchronization approach, enabling rapid sparse matrix assembly.

Speech Enhancement Algorithm Based on Microphone Array and Lightweight CRN for Hearing Aid

To address the performance and computational complexity issues in speech enhancement for hearing aids, a speech enhancement algorithm based on a microphone array and a lightweight two-stage convolutional recurrent network (CRN) is proposed. The algorithm consists of two main modules: a beamforming module and a post-filtering module. The beamforming module utilizes directional features and a complex time-frequency long short-term memory (CFT-LSTM) network to extract local representations and perform spatial filtering. The post-filtering module uses analogous encoding and two symmetric decoding structures, with stacked CFT-LSTM blocks in between. It further reduces residual noise and improves filtering performance by passing spatial information through an inter-channel masking module. Experimental results show that this algorithm outperforms existing methods on the generated hearing aid dataset and the CHIME-3 dataset, with fewer parameters and lower model complexity, making it suitable for hearing aid scenarios with limited computational resources.

Lateralization of memory function in temporal lobe epilepsy using scene memory fMRI.

Functional magnetic resonance imaging (fMRI) offers an alternative to the traditional Wada test for presurgical language and memory lateralization that carries almost no risk. However, fMRI lateralization of episodic memory remains challenging because the hippocampus, which is fundamental to episodic memory, is smaller, more prone to susceptibility artifact, and harder to functionally modulate than language regions. We previously showed that a complex scene memory task can lateralize memory function in the mesial temporal lobe. Using data acquired from N = 45 patients with temporal lobe epilepsy acquired with an improved stimulus paradigm and high-resolution fMRI, we now demonstrate that memory activation can be successfully lateralized within hippocampus proper. PLAIN LANGUAGE SUMMARY: Epilepsy surgery can improve seizure control in patients with temporal lobe epilepsy (TLE) that cannot be controlled with medications also, but ablation or removal of temporal lobe brain tissue can also cause cognitive deficits. Functional MRI (fMRI) can noninvasively map brain activation and perform well in lateralizing and localizing language function, but localizing and lateralizing memory function is more challenging. Building upon prior work using complex scene encoding to map memory function, we demonstrate that the use of high-resolution fMRI along with an optimized task paradigm allows memory activation to be detected within the hippocampus. Because the hippocampus is both a common site of TLE and a key region underlying memory, this approach is expected to contribute to presurgical evaluation of TLE.

The ECP SICM project: Managing complex memory hierarchies for exascale applications

The Exascale Computing Project (ECP)’s Simplified Interface to Complex Memories (SICM) effort focuses on developing universal interfaces for discovering, managing, and sharing data across complex memory hierarchies. These facilitate the exploitation of emerging memory technologies and support precise control over their various trade-offs such as high-bandwidth versus low-latency, persistent versus ephemeral, high-capacity versus low-capacity, and near-CPU versus near-GPU. SICM comprises three interrelated components: a low-level interface, a high-level interface, and a persistent-heap interface. The low-level SICM interface is intended for system and run-time developers as well as expert application developers who prefer full control of the memory objects used within their application. The high-level SICM interface builds upon the low-level interface, employing application-level profiling and analysis to optimize data management for complex memory hierarchies. The persistent-heap interface provides applications with a persistent memory allocator that can allocate custom C++ data structures in both block-storage and byte-addressable persistent memories.

A Functional Magnetic Resonance Imaging Study of Phrase Structure and Subject Island Violations.

In principle, functional neuroimaging provides uniquely informative data in addressing linguistic questions, because it can indicate distinct processes that are not apparent from behavioral data alone. This could involve adjudicating the source of unacceptability via the different patterns of elicited brain responses to different ungrammatical sentence types. However, it is difficult to interpret brain activations to syntactic violations. Such responses could reflect processes that have nothing intrinsically related to linguistic representations, such as domain-general executive function abilities. To facilitate the potential use of functional neuroimaging methods to identify the source of different syntactic violations, we conducted a functional magnetic resonance imaging experiment to identify the brain activation maps associated with two distinct syntactic violation types: phrase structure (created by inverting the order of two adjacent words within a sentence) and subject islands (created by extracting a wh-phrase out of an embedded subject). The comparison of these violations to control sentences surprisingly showed no indication of a generalized violation response, with almost completely divergent activation patterns. Phrase structure violations seemingly activated regions previously implicated in verbal working memory and structural complexity in sentence processing, whereas the subject islands appeared to activate regions previously implicated in conceptual-semantic processing, broadly defined. We review our findings in the context of previous research on syntactic and semantic violations using ERPs. Although our results suggest potentially distinct underlying mechanisms underlying phrase structure and subject island violations, our results are tentative and suggest important methodological considerations for future research in this area.

A Supervised Ml Algorithm for Detecting and Predicting Fraud Credit Card Transactions

A Credit card fraud presents an escalating threat in today's digital economy, leading to significant financial losses for consumers, businesses, and financial institutions. Traditional detection methods are increasingly inadequate due to the rise in online transactions and the evolving complexity of fraudulent activities. This research aims to create a reliable supervised machine learning model designed to detect and predict fraudulent credit card transactions in real-time. Our approach leverages advanced algorithms and extensive datasets to enhance transaction security, thereby mitigating financial risks. Current systems, like FICO's Fraud Detection System (FDS), utilize a range of supervised learning techniques, including Decision Trees, Random Forests, and Neural Networks. However, they face limitations, including data quality issues, high false positive rates, and the need for continuous retraining to adapt to evolving fraud patterns. To address these challenges, we propose an innovative anomaly detection method based on the Isolation Forest algorithm. Unlike traditional methods that rely on distance and density measures, Isolation Forests isolate anomalies by randomly selecting features and split values, making it more efficient in identifying fraudulent transactions, especially in imbalanced datasets. The proposed system boasts low linear time complexity and minimal memory requirements, enabling the construction of high-performing models even with large datasets. The research demonstrates that the Isolation Forest approach significantly outperforms traditional models in detecting credit card fraud, offering a promising solution for enhancing security in the financial ecosystem.

Nem az a kérdés, hogy az emlék téves-e, hanem hogy mennyire

Aim: The aim of the study is to investigate the background and psychology of a problem that often occurs in the application of law and is therefore of outstanding theoretical and practical importance – the wrong verdict due to the false memory of witnesses – in order to avoid the biggest judicial failures (justizmord). Methodology: The study was prepared by analysing international and Hungarian psychological theses and studies, describing relevant legal cases, and processing the international and domestic literature. Findings: The complexity, fragility and influence of human memory must be kept in mind during the procedure of the participants of the justice system, as well as the interrogation techniques to promote adequate memory, so that false decisions cannot be made in the justice system. Value: The credibility and accuracy of testimonies can be the key to finding out the material truth in the justice system, therefore the experiments described in the study and the techniques that support the testified testimony provide assistance to law enforcement.

Efficient memristor accelerator for transformer self-attention functionality

The adoption of transformer networks has experienced a notable surge in various AI applications. However, the increased computational complexity, stemming primarily from the self-attention mechanism, parallels the manner in which convolution operations constrain the capabilities and speed of convolutional neural networks (CNNs). The self-attention algorithm, specifically the matrix-matrix multiplication (MatMul) operations, demands a substantial amount of memory and computational complexity, thereby restricting the overall performance of the transformer. This paper introduces an efficient hardware accelerator for the transformer network, leveraging memristor-based in-memory computing. The design targets the memory bottleneck associated with MatMul operations in the self-attention process, utilizing approximate analog computation and the highly parallel computations facilitated by the memristor crossbar architecture. Remarkably, this approach resulted in a reduction of approximately 10 times in the number of multiply-accumulate (MAC) operations in transformer networks, while maintaining 95.47% accuracy for the MNIST dataset, as validated by a comprehensive circuit simulator employing NeuroSim 3.0. Simulation outcomes indicate an area utilization of 6895.7 μm2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mu m^2$$\\end{document}, a latency of 15.52 seconds, an energy consumption of 3 mJ, and a leakage power of 59.55 μW\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mu W$$\\end{document}. The methodology outlined in this paper represents a substantial stride towards a hardware-friendly transformer architecture for edge devices, poised to achieve real-time performance.

The hippocampal CA2 region discriminates social threat from social safety.

The dorsal cornu ammonis 2 (dCA2) region of the hippocampus enables the discrimination of novel from familiar conspecifics. However, the neural bases for more complex social-spatial episodic memories are unknown. Here we report that the spatial and social contents of an aversive social experience require distinct hippocampal regions. While dorsal CA1 (dCA1) pyramidal neurons mediate the memory of an aversive location, dCA2 pyramidal neurons enable the discrimination of threat-associated (CS+) from safety-associated (CS-) conspecifics in both female and male mice. Silencing dCA2 during encoding or recall trials disrupted social fear discrimination memory, resulting in fear responses toward both the CS+ and CS- mice. Calcium imaging revealed that the aversive experience strengthened and stabilized dCA2 representations of both the CS+ and CS- mice, with the incorporation of an abstract representation of social valence into representations of social identity. Thus, dCA2 contributes to both social novelty detection and the adaptive discrimination of threat-associated from safety-associated individuals during an aversive social episodic experience.

Investigating the Performance of the Informer Model for Streamflow Forecasting

Recent studies have shown the potential of transformer-based neural networks in increasing prediction capacity. However, classical transformers present several problems such as computational time complexity and high memory requirements, which make Long Sequence Time-Series Forecasting (LSTF) challenging. The contribution to the prediction of time series of flood events using deep learning techniques is examined, with a particular focus on evaluating the performance of the Informer model (a particular implementation of transformer architecture), which attempts to address the previous issues. The predictive capabilities of the Informer model are explored and compared to statistical methods, stochastic models and traditional deep neural networks. The accuracy, efficiency as well as the limits of the approaches are demonstrated via numerical benchmarks relating to real river streamflow applications. Using daily flow data from the River Test in England as the main case study, we conduct a rigorous evaluation of the Informer efficacy in capturing the complex temporal dependencies inherent in streamflow time series. The analysis is extended to encompass diverse time series datasets from various locations (>100) in the United Kingdom, providing insights into the generalizability of the Informer. The results highlight the superiority of the Informer model over established forecasting methods, especially regarding the LSTF problem. For a forecast horizon of 168 days, the Informer model achieves an NSE of 0.8 and maintains a MAPE below 10%, while the second-best model (LSTM) only achieves −0.63 and 25%, respectively. Furthermore, it is observed that the dependence structure of time series, as expressed by the climacogram, affects the performance of the Informer network.

An fMRI study of phrase structure and subject island violations.

In principle, functional neuroimaging provides uniquely informative data in addressing linguistic questions, because it can indicate distinct processes that are not apparent from behavioral data alone. This could involve adjudicating the source of unacceptability via the different patterns of elicited brain responses to different ungrammatical sentence types. However, it is difficult to interpret brain activations to syntactic violations. Such responses could reflect processes that have nothing intrinsically related to linguistic representations, such as domain-general executive function abilities. In order to facilitate the potential use of functional neuroimaging methods to identify the source of different syntactic violations, we conducted an fMRI experiment to identify the brain activation maps associated with two distinct syntactic violation types: phrase structure (created by inverting the order of two adjacent words within a sentence) and subject islands (created by extracting a wh-phrase out of an embedded subject). The comparison of these violations to control sentences surprisingly showed no indication of a generalized violation response, with almost completely divergent activation patterns. Phrase structure violations seemingly activated regions previously implicated in verbal working memory and structural complexity in sentence processing, whereas the subject islands appeared to activate regions previously implicated in conceptual-semantic processing, broadly defined. We review our findings in the context of previous research on syntactic and semantic violations using event-related potentials. Although our results suggest potentially distinct underlying mechanisms underlying phrase structure and subject island violations, our results are tentative and suggest important methodological considerations for future research in this area.

SparseAuto: An Auto-scheduler for Sparse Tensor Computations using Recursive Loop Nest Restructuring

Automated code generation and performance enhancements for sparse tensor algebra have become essential in many real-world applications, such as quantum computing, physical simulations, computational chemistry, and machine learning. General sparse tensor algebra compilers are not always versatile enough to generate asymptotically optimal code for sparse tensor contractions. This paper shows how to generate asymptotically better schedules for complex sparse tensor expressions using kernel fission and fusion. We present generalized loop restructuring transformations to reduce asymptotic time complexity and memory footprint. Furthermore, we present an auto-scheduler that uses a partially ordered set (poset)-based cost model that uses both time and auxiliary memory complexities to prune the search space of schedules. In addition, we highlight the use of Satisfiability Module Theory (SMT) solvers in sparse auto-schedulers to approximate the Pareto frontier of better schedules to the smallest number of possible schedules, with user-defined constraints available at compile-time. Finally, we show that our auto-scheduler can select better-performing schedules and generate code for them. Our results show that the auto-scheduler provided schedules achieve orders-of-magnitude speedup compared to the code generated by the Tensor Algebra Compiler (TACO) for several computations on different real-world tensors.

A Note on Related-Tweakey Impossible Differential Attacks

In this note we review the technique proposed at ToSC 2018 by Sadeghi et al. for attacks built upon several related-tweakey impossible differential trails. We show that the initial encryption queries are improper and lead the authors to misevaluate a filtering value in the key recovery phase. We identified 4 other papers (from Eurocrypt, DCC, and 2 from ToSC) that follow on the results of Sadeghi et al. and in three of them the flawed technique was reused. We thus present a careful analysis of these types of attacks and give generic complexity formulas similar to the ones proposed by Boura et al. at Asiacrypt 2014. We apply these to the aforementioned papers and provide patched versions of their attacks. The main consequence is an increase in the memory complexity. We show that in many cases (a notable exception being quantum impossible differentials) it is possible to recover the numeric time estimates of the flawed analysis, and in all cases we were able to build a correct attack reaching the same number of rounds.

AI As Artificial Memory: A Global Reconfiguration of Our Collective Memory Practices?

Abstract ai’s evolution has drastically reshaped our relationship with collective memory. Large ai models, like gpt and dall·E, have emerged as memory agents, assimilating vast cultural and historical data across centuries. Unlike traditional digital repositories, these tools employ an analogical approach, capturing the complexity of human memory. Their potential to weave together fragmented pieces of the past offers new ways to delve into historical and cultural data. Tech giants are ambitiously positioning these ai models as primary mediators for information access, integrating them into popular platforms. While they present new forms of collective memory, their widespread use raises profound ethical concerns. Challenges arise regarding historical truth, representativeness, and data reuse, leading to debates on data ownership and the governance of digital collective memory.

The effects of multitalker babble noise on children's cognitive performance

Classroom noise, particularly irrelevant speech or background babble noise, can have detrimental effects on academic tasks and learning more generally. Researchers have proposed two alternative explanations of the interference of babble noise on non-auditory academic tasks, such as reading or math: i) Babble noise may generate interference on these tasks due to the automatic access of speech sounds (babbling) to phonological short-term memory; or ii) by capturing children's attention, thereby diverting general attentional resources from the task. To test these hypotheses, we asked 3rd to 5th graders to perform two visual attention and two verbal working memory tasks (digit span and reading span) in both quiet and noisy conditions. Multitalker babble noise was administered through headphones at 65dB. The results revealed significant noise effects only in the most complex verbal working memory task (reading span). Interestingly, a significant difference between conditions was also observed in one of the visual attention tasks, where higher scores were, however, achieved in the presence of noise. These preliminary findings will be discussed in light of the current literature.

Towards a Generalized Bayesian Model of Reconstructive Memory

AbstractPrior knowledge has long been known to shape reconstruction from memory. An individual stimulus from a category is often remembered to be closer to the center of that category than its true location. This effect, together with more complex memory effects that involve prior knowledge at multiple levels of abstraction, has been successfully explained by the Category Adjustment Model (CAM; Huttenlocher et al. Journal of Experimental Psychology: General, 129(2), 220, 2000) and its extensions. However, recent experimental results diverge from CAM’s predictions showing that reconstructive memory for atypical category examples is influenced by the category center less than that of typical category examples. To unify these findings, we propose a generalized Bayesian model of reconstructive memory, called the generalized CAM model (g-CAM). We demonstrate through simulations that g-CAM can account for previously known effects of reconstructive memory, while additionally capturing recent empirical findings involving atypical category examples.

Multidimensional zero-correlation linear cryptanalysis on uBlock and RAIN

uBlock and RAIN are both SPN structure lightweight block ciphers with sufficient security against traditional attacks. This article mainly focuses on the zero-correlation linear cryptanalysis of uBlock-128/128 and RAIN-64 without considering the whitening key for the first time. On the one hand, combining the linear mask propagation rules of S-box with matrix method, a large number of 4-round zero-correlation linear approximations for uBlock-128/128 are obtained. Therefore, a 6-round key recovery attack is carried out with partial-compression technique, which can recover 48-bit subkeys with data complexity 2 126.41 known plaintexts, time complexity 2 122.82 times of 6-round of algorithm encryptions, and memory complexity 2 48 nibbles. On the other hand, a similar method is used on RAIN-64, thus a 10-round attack is performed, which recovers 64-bit subkeys with data complexity 2 61.95 known plaintexts, time complexity 2 74.26 times of 10-round algorithm encryptions, and memory complexity 2 64 nibbles. The results show that both uBlock and RAIN are safe enough to resist zero-correlation linear cryptanalysis.

CvFormer: Cross-view transFormers with pre-training for fMRI analysis of human brain

In recent years, functional magnetic resonance imaging (fMRI) has been widely utilized to diagnose neurological disease, by exploiting the region of interest (RoI) nodes as well as their connectivities in human brain. However, most of existing works only rely on either RoIs or connectivities, neglecting the potential for complementary information between them. To address this issue, we study how to discover the rich cross-view information in fMRI data of human brain. This paper presents a novel method for cross-view analysis of fMRI data of the human brain, called Cross-view transFormers (CvFormer). CvFormer employs RoI and connectivity encoder modules to generate two separate views of the human brain, represented as RoI and sub-connectivity tokens. Then, basic transformer modules can be used to process the RoI and sub-connectivity tokens, and cross-view modules integrate the complement information across two views. Furthermore, CvFormer uses a global token for each branch as a query to exchange information with other branches in cross-view modules, which only requires linear time for both computational and memory complexity instead of quadratic time. To enhance the robustness of the proposed CvFormer, we propose a two-stage strategy to train its parameters. To be specific, RoI and connectivity views can be firstly utilized as self-supervised information to pre-train the CvFormer by combining it with contrastive learning and then fused to finetune the CvFormer using label information. Experiment results on two public ABIDE and ADNI datasets can show clear improvements by the proposed CvFormer, which can validate its effectiveness and superiority.

Multilevel approximation of Gaussian random fields: Covariance compression, estimation, and spatial prediction

The distribution of centered Gaussian random fields (GRFs) indexed by compacta such as smooth, bounded Euclidean domains or smooth, compact and orientable manifolds is determined by their covariance operators. We consider centered GRFs given as variational solutions to coloring operator equations driven by spatial white noise, with an elliptic self-adjoint pseudodifferential coloring operator from the Hörmander class. This includes the Matérn class of GRFs as a special case. Using biorthogonal multiresolution analyses on the manifold, we prove that the precision and covariance operators, respectively, may be identified with bi-infinite matrices and finite sections may be diagonally preconditioned rendering the condition number independent of the dimension p of this section. We prove that a tapering strategy by thresholding applied on finite sections of the bi-infinite precision and covariance matrices results in optimally numerically sparse approximations. That is, asymptotically only linearly many nonzero matrix entries are sufficient to approximate the original section of the bi-infinite covariance or precision matrix using this tapering strategy to arbitrary precision. The locations of these nonzero matrix entries can be determined a priori. The tapered covariance or precision matrices may also be optimally diagonally preconditioned. Analysis of the relative size of the entries of the tapered covariance matrices motivates novel, multilevel Monte Carlo (MLMC) oracles for covariance estimation, in sample complexity that scales log-linearly with respect to the number p of parameters. In addition, we propose and analyze novel compressive algorithms for simulating and kriging of GRFs. The complexity (work and memory vs. accuracy) of these three algorithms scales near-optimally in terms of the number of parameters p of the sample-wise approximation of the GRF in Sobolev scales.

Methods for calculating gliding-box lacunarity efficiently on large datasets

Lacunarity has proven to be a useful, multifaceted tool for image analysis in several different scientific fields, from geography to virology, which has lent increasing importance to the lacunarity analysis of large datasets. It can be most reliably calculated with the so-called gliding-box method, but the evaluation process can be exceedingly time-consuming and unviable as this algorithm is not designed to operate on large datasets. Here we introduce two novel methods that can calculate gliding-box lacunarity orders of magnitude faster than the original method without any loss of accuracy. We compare these methods with the original as well as with two already existing optimized methods based on runtime memory usage and complexity. The application of all five methods for both 2D and 3D datasets analysis confirms that each of the four optimized methods are orders of magnitude faster than the original one, but each has its advantages and limitations.