32584 research outputs found
Sort by
Transformers as Transducers
Get PDFWe study the sequence-to-sequence mapping capacity of transformers by relating them to finite transducers, and find that they can express surprisingly large classes of (total functional) transductions. We do so using variants of RASP, a programming language designed to help people “think like transformers,” as an intermediate representation. We extend the existing Boolean variant B-RASP to sequence-to-sequence transductions and show that it computes exactly the first-order rational transductions (such as string rotation). Then, we introduce two new extensions. B-RASP[pos] enables calculations on positions (such as copying the first half of a string) and contains all first-order regular transductions. S-RASP adds prefix sum, which enables additional arithmetic operations (such as squaring a string) and contains all first-order polyregular transductions. Finally, we show that masked average-hard attention transformers can simulate S-RASP
Exploration of the effects of Tai Chi practice on lower limb corticomuscular coherence during balance-demanding virtual reality conditions in older adults
Get PDFTai Chi practice has been widely adopted to improve balance and prevent falls in older adults. However, the neural mechanisms underlying the benefits of Tai Chi are difficult to evaluate during traditional balance assessments. The goal of this study was to evaluate the effects of Tai Chi and healthy aging on corticomuscular coherence (CMC) while standing in virtual balance-demanding environments. We recorded neural, muscular, and behavioral data in an immersive virtual reality environment while implementing sensory and mechanical perturbations to introduce high postural threats. Through the acquisition of electroencephalography and electromyography signals, we examined β and γ CMC changes in frontal, central, parietal, and occipital cortical areas and ankle plantar- and dorsi-flexors in older adults (n = 10), older adults with Tai Chi practice (n = 10), and young adults (n = 10). The results showed that older adults have higher γ CMC in comparison with Tai Chi practitioners and young adults as evaluated by the magnitude square coherence. Increased β and γ CMC correlated with decreased mediolateral postural sway in older adults, while young adults demonstrated the opposite relationship. Furthermore, lower tibialis anterior and soleus β CMC were found in older adults during ground conditions compared to Tai Chi practitioners and young adults. The results demonstrate the effects of aging and Tai Chi on CMC during balance-demanding standing tasks, and the potential application of the novel system to quantify cortical and muscular adaptation after rehabilitation
Generating Synthetic Malware Samples Using Generative AI
Get PDFMalware attacks have a significant negative impact on organizations of varied scales in the field of cybersecurity. Recently, malware researchers have increasingly turned to machine learning techniques to combat sophisticated obfuscation methods used in malware. However, collecting a diverse set of malware samples with various obfuscation techniques is challenging and often takes years, especially for newly developed malware. This issue is further compounded by a well-known limitation of machine learning models: their poor performance when training data is scarce. In this paper, we propose a new system for generating synthetic malware samples to augment imbalanced malware dataset. Our approach decomposes malware binary samples into mnemonic opcode sequences, leveraging natural language processing to extract contextual meaning behind malware opcode features to aid the learning of generative AI (GenAI) employed in this paper, Generative Adversarial Networks (GAN), Wasserstein Generative Adversarial Networks with Gradient Penalty (WGAN-GP), and a modified Diffusion model. The experiment results show that augmenting training data with Diffusion-based synthetic data significantly improves classification performance for minor classes by up to 60% on average. This enhancement ultimately leads to an overall malware classification performance of 96%, an 8% improvement. These findings demonstrate the high quality and fidelity of the synthetic data, its robustness, and its potential applications in malware analysis. Specifically, synthetic malware data proves effective in improving the classification of minor malware classes and detection rates, even though the size of known malware data is significantly small
Spartan Daily, September 24, 2025
Get PDFVolume 165, Issue 12https://scholarworks.sjsu.edu/spartan_daily_2025/1056/thumbnail.jp
Comparative Analysis of Mineral Content in Plant-based Milk Alternatives: Implications for Renal Patient Education
Get PDFThe plant-based milk alternative (PBMA) market is expanding rapidly, offering a diverse range of options. However, the mineral composition of PBMA varies significantly across brands and milk types, posing challenges in educating individuals with renal disease. This study aimed to quantify and compare key minerals of concern for renal patients in commercially available PBMA using inductively coupled plasma optical emission spectrometry.
PBMA products available in the San Francisco Bay Area were purchased from chain grocery stores in spring 2023 and early January 2024. A total of 134 unique PBMA products from 28 brands and 11 different milk-alternative sources were analyzed. These were categorized into four groups: Legume (pea, soy), Nut (almond, walnut, cashew, macadamia, coconut), Seed (flax, hemp), and Cereal (oat, rice).
Mean calcium, potassium, phosphorus, and sodium content varied across groups. Legume milk had the highest mean calcium (348.37±175.63 mg) and potassium (381.09±76.51 mg) per 240 mL Cereal milk had the lowest mean calcium (202.09±154.52 mg), while Seed milk had the lowest potassium (61.39±67.90 mg). Nut milk had the lowest phosphorus (62.31±62.48 mg). Legume milk had the lowest sodium (96.29±44.19 mg), whereas Seed milk had the highest (151.92±31.25 mg). The mean potassium and phosphorus content in Legume milk exceeded renal dietary recommendations (\u3c 200 mg potassium, \u3c 150 mg phosphorus per serving) and individuals with chronic kidney disease should limit or monitor its intake.
Given the variability in mineral content, renal patients may benefit from individualized PBMA guidance to optimize their health outcomes
The Anti-Work Attitude Scale: Conceptualizing and Measuring the Resistance to the Institution of Work
Get PDFRecent shifts in labor discourse have drawn attention to rising burnout, declining organizational commitment, and diminishing job satisfaction, which challenge traditional assumptions about the centrality and desirability of work. These tensions have given rise to the concept of anti-work, a growing discourse that questions the moral, economic, and psychological foundations of work itself. Despite its cultural relevance, anti-work remains largely undertheorized and unmeasured in empirical research. Drawing from interdisciplinary scholarship, this study defined anti-work as a negative critique of work as a social construct, questioning its value, purpose, and impact on individuals and society. Anti-work was operationalized through two higher-order dimensions, systemic critique of work and individual experience at work, composed of theoretically grounded subdimensions such as meaningful work, skepticism, worker exploitation, will to work, and employee engagement. The purpose of this study was to develop and preliminarily validate a scale, the Anti-Work Attitude Scale (AWAS) that captures anti-work attitudes from both systemic and individual perspectives. The AWAS, consisting of 42 items (6 items for each of seven subdimensions), exhibited strong internal consistency. Future research directions, including the need to assess the dimensionality and validity of the AWAS, are discussed
Could Password Sharing Entitle You to Monitor Your Partner\u27s Social Media Accounts?
Get PDFSocial media offers a plethora of strangers cum friends whose photo shopped images seem idealistic and more appealing than one\u27s primary partner. Due to the physical absence and lack of non-verbal cues, online communications can quickly become aggressive and emotionally intimate. Emotional intimacy between online friends is considered a betrayal in a dyadic relationship. To protect mates from external relationship threats, romantic partners may request and/or coerce their significant other for social media account passwords. In a sample of 299 adults (women = 246) between18 and 72 years (M = 21.12, SD = 5.39), we explored the participants’ attitudes toward password sharing, their password sharing behavior with the significant other, and finally seeking social media account passwords from their significant other. We also tested if sharing one\u27s passwords (password sharing behavior) would mediate the relationship between the participant\u27s password sharing attitude and the significant other\u27s account monitoring. We found that favorable password sharing attitude may predict password sharing behavior and the significant other\u27s SNS account monitoring. Password sharing behavior also partially mediated the relationship between password sharing attitude and the significant other\u27s account monitoring. Implications and limitations of the study are discussed
Biopsychosocial Contexts of Stress: From Mechanisms to Targeted Interventions Presentation
No full textIndividuals exposed to early life stress exhibit wide variability in outcomes—but what drives this difference? This talk explores how biopsychosocial mechanisms shape risk and resilience, examines group differences in stress responses, and discusses implications for interventions. Dr. Chong presents three studies that (1) investigate the combined effects of early life stress and stress responses on one’s outcomes, (2) review racial disparities in stress responses, and (3) evaluate the impact of mindfulness interventions on stress system coordination. Together, these findings emphasize the importance of a contextually grounded approach to developmental risk and resilience, informing more equitable clinical psychology strategies.https://scholarworks.sjsu.edu/uss/1066/thumbnail.jp
Spartan Daily, September 30, 2025
Get PDFVolume 165, Issue 16https://scholarworks.sjsu.edu/spartan_daily_2025/1059/thumbnail.jp
Computing Education in African Countries: A Literature Review and Contextualised Learning Materials
Get PDFThis report begins with a literature review of computing education in Africa. We found a substantial body of work, scattered over more than 80 venues, which we have brought together here for the first time. Several important themes emerge in this dataset, including the need to contextualise computing education. In the second part of this report we investigate contextualisation further. We present a pilot study, grounded in the literature review, of the development of course materials, sample code, and programming assignments for introductory programming, contextualised for six African countries: Botswana, Egypt, Ghana, Nigeria, South Africa, and Zambia. We include the materials, report on a preliminary evaluation of the materials by fellow educators in African countries, and suggest a process by which other educators could develop materials for their local contexts