Physical reservoir computing (PRC) is a promising brain-inspired computing architecture for overcoming the von Neumann bottleneck by utilizing the intrinsic dynamics of physical systems. However, a major obstacle to its real-world implementation lies in the tension between extracting sufficient information for high computational performance and maintaining a hardware-feasible, high-speed architecture. Here, we report spectral dynamics reservoir computing (SDRC), a broadly applicable framework based on analogue filtering and envelope detection that bridges this gap. SDRC effectively exploits the fast spectral dynamics embedded in short-time, coarse spectra of material responses to attain strong computational capability while maintaining high-speed processing and minimal hardware overhead. This approach circumvents the need for implementation-intensive, precision-sensitive integrated circuits required in high-speed time-multiplexing measurements, while enabling real-time use of the material's spectral manifold as a high-dimensional computational resource. We implement and experimentally demonstrate SDRC applied to spin waves that achieves state-of-the-art-level performance with only 56 nodes on benchmark tasks of parity-check and second-order nonlinear autoregressive moving average, as well as high accuracy of 98.0% on a real-world problem of speech recognition.
Fully Homomorphic Encryption (FHE) is rapidly emerging as a promising foundation for privacy-preserving cloud services, enabling computation directly on encrypted data. As FHE implementations mature and begin moving toward practical deployment in domains such as secure finance, biomedical analytics, and privacy-preserving AI, a critical question remains insufficiently explored: how reliable is FHE computation on real hardware? This question is especially important because, compared with plaintext computation, FHE incurs much higher computational overhead, making it more susceptible to transient hardware faults. Moreover, data corruptions are likely to remain silent: the FHE service has no access to the underlying plaintext, causing unawareness even though the corresponding decrypted result has already been corrupted. To this end, we conduct a comprehensive evaluation of SDCs in FHE ciphertext computation. Through large-scale fault-injection experiments, we characterize the vulnerability of FHE to transient faults, and through a theoretical analysis of error-propagation behaviors, we gain deeper algorithmic insight into the mechanisms underlying this vulnerability. We further assess the effectiveness of different fault-tolerance mechanisms for mitigating these faults.
Atousa Jafari, Mahdi Taheri, Hassan Ghasemzadeh Mohammadi
et al.
This paper presents a compression framework for Reservoir Computing that enables systematic design-space exploration of trade-offs among quantization levels, pruning rates, model accuracy, and hardware efficiency. The proposed approach leverages a sensitivity-based pruning mechanism to identify and remove less critical quantized weights with minimal impact on model accuracy, thereby reducing computational overhead while preserving accuracy. We perform an extensive trade-off analysis to validate the effectiveness of the proposed framework and the impact of pruning and quantization on model performance and hardware parameters. For this evaluation, we employ three time-series datasets, including both classification and regression tasks. Experimental results across selected benchmarks demonstrate that our proposed approach maintains high accuracy while substantially improving computational and resource efficiency in FPGA-based implementations, with variations observed across different configurations and time series applications. For instance, for the MELBOEN dataset, an accelerator quantized to 4-bit at a 15\% pruning rate reduces resource utilization by 1.2\% and the Power Delay Product (PDP) by 50.8\% compared to an unpruned model, without any noticeable degradation in accuracy.
The integration of Large Language Models (LLMs) into computer applications has introduced transformative capabilities but also significant security challenges. Existing safety alignments, which primarily focus on semantic interpretation, leave LLMs vulnerable to attacks that use non-standard data representations. This paper introduces ArtPerception, a novel black-box jailbreak framework that strategically leverages ASCII art to bypass the security measures of state-of-the-art (SOTA) LLMs. Unlike prior methods that rely on iterative, brute-force attacks, ArtPerception introduces a systematic, two-phase methodology. Phase 1 conducts a one-time, model-specific pre-test to empirically determine the optimal parameters for ASCII art recognition. Phase 2 leverages these insights to launch a highly efficient, one-shot malicious jailbreak attack. We propose a Modified Levenshtein Distance (MLD) metric for a more nuanced evaluation of an LLM's recognition capability. Through comprehensive experiments on four SOTA open-source LLMs, we demonstrate superior jailbreak performance. We further validate our framework's real-world relevance by showing its successful transferability to leading commercial models, including GPT-4o, Claude Sonnet 3.7, and DeepSeek-V3, and by conducting a rigorous effectiveness analysis against potential defenses such as LLaMA Guard and Azure's content filters. Our findings underscore that true LLM security requires defending against a multi-modal space of interpretations, even within text-only inputs, and highlight the effectiveness of strategic, reconnaissance-based attacks. Content Warning: This paper includes potentially harmful and offensive model outputs.
The growth of cloud computing has revolutionized data processing and storage capacities to another levels of scalability and flexibility. But in the process, it has created a huge challenge of security, especially in terms of safeguarding sensitive data. Classical security practices, including encryption at rest and during transit, fail to protect data in use and expose it to various possible breaches. In response to this problem , Confidential Computing has been a tool ,seeking to secure data in processing by usage of hardware-based Trusted Execution Environments (TEEs). TEEs, including Intel's Software Guard Extensions (SGX) and ARM's TrustZone, offers protected contexts within the processor, where data is kept confidential ,intact and secure , even with malicious software or compromised operating systems. In this research, we have explored the architecture and security features of TEEs like Intel SGX and ARM TrustZone, and their effectiveness in improving cloud data security. From a thorough literature survey ,we have analyzed the deployment strategies, performance indicators, and practical uses of these TEEs for the same purpose. In addition, we have discussed the issues regarding deployment, possible weaknesses, scalability issues, and integration issues. Our results focuses on the central position of TEEs in strengthening and advancing cloud security infrastructures, pointing towards their ability to create a secure foundation for Confidential Computing.
Evelyne Ringoot, Rabab Alomairy, Valentin Churavy
et al.
This paper presents a portable, GPU-accelerated implementation of a QR-based singular value computation algorithm in Julia. The singular value ecomposition (SVD) is a fundamental numerical tool in scientific computing and machine learning, providing optimal low-rank matrix approximations. Its importance has increased even more in large-scale machine learning pipelines, including large language models (LLMs), where it enables low-rank adaptation (LoRA). The implemented algorithm is based on the classic two-stage QR reduction, consisting of successive matrix reduction to band form and bidiagonal form. Our implementation leverages Julia's multiple dispatch and metaprogramming capabilities, integrating with the GPUArrays and KernelAbstractions frameworks to provide a unified type and hardware-agnostic function. It supports diverse GPU architectures and data types, and is, to our knowledge, the first GPU-accelerated singular value implementation to support Apple Metal GPUs and half precision. Performance results on multiple GPU backends and data types demonstrate that portability does not require sacrificing performance: the unified function outperforms most linear algebra libraries (MAGMA, SLATE, rocSOLVER, oneMKL) for matrix sizes larger than 1024x1024, and achieves 80%-90% of the performance of cuSOLVER for large matrices.
Paris Avgeriou, Nauman bin Ali, Marcos Kalinowski
et al.
Increasingly, courses on Empirical Software Engineering research methods are being offered in higher education institutes across the world, mostly at the M.Sc. and Ph.D. levels. While the need for such courses is evident and in line with modern software engineering curricula, educators designing and implementing such courses have so far been reinventing the wheel; every course is designed from scratch with little to no reuse of ideas or content across the community. Due to the nature of the topic, it is rather difficult to get it right the first time when defining the learning objectives, selecting the material, compiling a reader, and, more importantly, designing relevant and appropriate practical work. This leads to substantial effort (through numerous iterations) and poses risks to the course quality. This chapter attempts to support educators in the first and most crucial step in their course design: creating the syllabus. It does so by consolidating the collective experience of the authors as well as of members of the Empirical Software Engineering community; the latter was mined through two working sessions and an online survey. Specifically, it offers a list of the fundamental building blocks for a syllabus, namely course aims, course topics, and practical assignments. The course topics are also linked to the subsequent chapters of this book, so that readers can dig deeper into those chapters and get support on teaching specific research methods or cross-cutting topics. Finally, we guide educators on how to take these building blocks as a starting point and consider a number of relevant aspects to design a syllabus to meet the needs of their own program, students, and curriculum.
Josep Lopez Camunas, Cristina Bustos, Yanjun Zhu
et al.
Understanding emotional signals in older adults is crucial for designing virtual assistants that support their well-being. However, existing affective computing models often face significant limitations: (1) limited availability of datasets representing older adults, especially in non-English-speaking populations, and (2) poor generalization of models trained on younger or homogeneous demographics. To address these gaps, this study evaluates state-of-the-art affective computing models -- including facial expression recognition, text sentiment analysis, and smile detection -- using videos of older adults interacting with either a person or a virtual avatar. As part of this effort, we introduce a novel dataset featuring Spanish-speaking older adults engaged in human-to-human video interviews. Through three comprehensive analyses, we investigate (1) the alignment between human-annotated labels and automatic model outputs, (2) the relationships between model outputs across different modalities, and (3) individual variations in emotional signals. Using both the Wizard of Oz (WoZ) dataset and our newly collected dataset, we uncover limited agreement between human annotations and model predictions, weak consistency across modalities, and significant variability among individuals. These findings highlight the shortcomings of generalized emotion perception models and emphasize the need of incorporating personal variability and cultural nuances into future systems.
Sonja Hyrynsalmi, Ella Peltonen, Fanny Vainionpää
et al.
In the extant literature, there has been discussion on the drivers and motivations of minorities to enter the software industry. For example, universities have invested in more diverse imagery for years to attract a more diverse pool of students. However, in our research, we consider whether we understand why students choose their current major and how they did in the beginning decided to apply to study software engineering. We were also interested in learning if there could be some signs that would help us in marketing to get more women into tech. We approached the topic via an online survey (N = 78) sent to the university students of software engineering in Finland. Our results show that, on average, women apply later to software engineering studies than men, with statistically significant differences between genders. Additionally, we found that marketing actions have different impacts based on gender: personal guidance in live events or platforms is most influential for women, whereas teachers and social media have a more significant impact on men. The results also indicate two main paths into the field: the traditional linear educational pathway and the adult career change pathway, each significantly varying by gender
Francisco Chicano, Gabiel Luque, Zakaria Abdelmoiz Dahi
et al.
Quantum computers leverage the principles of quantum mechanics to do computation with a potential advantage over classical computers. While a single classical computer transforms one particular binary input into an output after applying one operator to the input, a quantum computer can apply the operator to a superposition of binary strings to provide a superposition of binary outputs, doing computation apparently in parallel. This feature allows quantum computers to speed up the computation compared to classical algorithms. Unsurprisingly, quantum algorithms have been proposed to solve optimization problems in quantum computers. Furthermore, a family of quantum machines called quantum annealers are specially designed to solve optimization problems. In this paper, we provide an introduction to quantum optimization from a practical point of view. We introduce the reader to the use of quantum annealers and quantum gate-based machines to solve optimization problems.
Sukhpal Singh Gill, Oktay Cetinkaya, Stefano Marrone
et al.
The recent development of quantum computing, which uses entanglement, superposition, and other quantum fundamental concepts, can provide substantial processing advantages over traditional computing. These quantum features help solve many complex problems that cannot be solved otherwise with conventional computing methods. These problems include modeling quantum mechanics, logistics, chemical-based advances, drug design, statistical science, sustainable energy, banking, reliable communication, and quantum chemical engineering. The last few years have witnessed remarkable progress in quantum software and algorithm creation and quantum hardware research, which has significantly advanced the prospect of realizing quantum computers. It would be helpful to have comprehensive literature research on this area to grasp the current status and find outstanding problems that require considerable attention from the research community working in the quantum computing industry. To better understand quantum computing, this paper examines the foundations and vision based on current research in this area. We discuss cutting-edge developments in quantum computer hardware advancement and subsequent advances in quantum cryptography, quantum software, and high-scalability quantum computers. Many potential challenges and exciting new trends for quantum technology research and development are highlighted in this paper for a broader debate.
Mateusz Kocot, Krzysztof Misan, Valentina Avati
et al.
Measurements from particle timing detectors are often affected by the time walk effect caused by statistical fluctuations in the charge deposited by passing particles. The constant fraction discriminator (CFD) algorithm is frequently used to mitigate this effect both in test setups and in running experiments, such as the CMS-PPS system at the CERN's LHC. The CFD is simple and effective but does not leverage all voltage samples in a time series. Its performance could be enhanced with deep neural networks, which are commonly used for time series analysis, including computing the particle arrival time. We evaluated various neural network architectures using data acquired at the test beam facility in the DESY-II synchrotron, where a precise MCP (MicroChannel Plate) detector was installed in addition to PPS diamond timing detectors. MCP measurements were used as a reference to train the networks and compare the results with the standard CFD method. Ultimately, we improved the timing precision by 8% to 23%, depending on the detector's readout channel. The best results were obtained using a UNet-based model, which outperformed classical convolutional networks and the multilayer perceptron.
Ravi Sahita, Atish Patra, Vedvyas Shanbhogue
et al.
Multi-tenant computing platforms are typically comprised of several software and hardware components including platform firmware, host operating system kernel, virtualization monitor, and the actual tenant payloads that run on them (typically in a virtual machine, container, or application). This model is well established in large scale commercial deployment, but the downside is that all platform components and operators are in the Trusted Computing Base (TCB) of the tenant. This aspect is ill-suited for privacy-oriented workloads that aim to minimize the TCB footprint. Confidential computing presents a good stepping-stone towards providing a quantifiable TCB for computing. Confidential computing [1] requires the use of a HW-attested Trusted Execution Environments for data-in-use protection. The RISC-V architecture presents a strong foundation for meeting the requirements for Confidential Computing and other security paradigms in a clean slate manner. This paper describes a reference architecture and discusses ISA, non-ISA and system-on-chip (SoC) requirements for confidential computing on RISC-V Platforms. It discusses proposed ISA and non-ISA Extension for Confidential Virtual Machine for RISC-V platforms, referred to as CoVE.
Analytics corresponds to a relevant and challenging phase of Big Data. The generation of knowledge from extensive data sets (petabyte era) of varying types, occurring at a speed able to serve decision makers, is practiced using multiple areas of knowledge, such as computing, statistics, data mining, among others. In the Big Data domain, Analytics is also considered as a process capable of adding value to the organizations. Besides the demonstration of value, Analytics should also consider operational tools and models to support decision making. To adding value, Analytics is also presented as part of some Big Data value chains, such the Information Value Chain presented by NIST among others, which are detailed in this article. As well, some maturity models are presented, since they represent important structures to favor continuous implementation of Analytics for Big Data, using specific technologies, techniques and methods. Hence, through an in-depth research, using specific literature references and use cases, we seeks to outline an approach to determine the Analytical Engineering for Big Data Analytics considering four pillars: Data, Models, Tools and People; and three process groups: Acquisition, Retention and Revision; in order to make feasible and to define an organization, possibly designated as an Analytics Organization, responsible for generating knowledge from the data in the field of Big Data Analytics.
Automated affective computing in the wild is a challenging task in the field of computer vision. This paper presents three neural network-based methods proposed for the task of facial affect estimation submitted to the First Affect-in-the-Wild challenge. These methods are based on Inception-ResNet modules redesigned specifically for the task of facial affect estimation. These methods are: Shallow Inception-ResNet, Deep Inception-ResNet, and Inception-ResNet with LSTMs. These networks extract facial features in different scales and simultaneously estimate both the valence and arousal in each frame. Root Mean Square Error (RMSE) rates of 0.4 and 0.3 are achieved for the valence and arousal respectively with corresponding Concordance Correlation Coefficient (CCC) rates of 0.04 and 0.29 using Deep Inception-ResNet method.