This work presents the experimental characterization and empirical modeling of the dielectric properties at microwave frequencies of two conductive liquid systems: de-ionized water/herbicide mixtures and lithium-salt electrolyte solutions, relevant to environmental/health monitoring and energy storage applications, respectively. Measurements are performed using the coaxial probe technique over a frequency range starting from 500 MHz and up to some tenths of GHz. The dielectric permittivity at various concentration levels is modeled using the single-pole Debye model with frequency-dependent conductivity component. The electrode polarization effect at the probe-sample interface is systematically analyzed, proposing a low-frequency measurement limit that shifts toward higher frequencies as the conductivity increases; a relative-deviation analysis on both permittivity components quantitatively assesses this limitation. Within the validated frequency range, the Debye formulation accurately reproduces the properties of both liquid systems, showing that conductive and interfacial effects dominate at the lower frequencies while dipolar relaxation prevails at higher frequencies.
Freelance workers must continually acquire new skills to remain competitive in online labor markets, yet they lack the organizational training, mentorship, and infrastructure available to traditional employees. Generative AI-powered tools like ChatGPT are reshaping market skill demands while also offering new forms of on-demand learning support to meet those demands. Despite growing interest in AI-powered learning tools, little is known about how freelancers actually use these tools to learn, the challenges they encounter, and how generative AI for learning interacts with precarity and competition in platform-based work. We present a mixed-methods study combining a survey and semi-structured interviews with freelance knowledge workers. Grounded in self-directed learning theory, we examine how freelancers integrate generative AI tools into their learning practices. Our findings show that freelancers increasingly rely on generative AI to structure learning and support exploratory skill acquisition, but do not treat it as their primary learning resource due to inconsistency, lack of contextual relevance, and verification overhead. We identify a shift from learning as growth to learning as survival, where upskilling is oriented toward immediate market viability rather than long-term development. We also surface a structural challenge we term invisible competencies, in which workers acquire skills through generative AI tools but lack credible ways to signal or validate these skills in competitive freelance markets. Based on these insights, we offer design recommendations for generative AI-powered learning tools for freelancers.
As debates over the impact of smartphones on young people intensify, we urgently need ways to identify which smartphone behaviours are related to poorer mental health. While many approaches focus on specific harmful or beneficial smartphone activities, here we instead suggest that the relationship between smartphones and mental health may partly depend on how users transition between activities. To test this, we develop a novel network analysis method to analyze objective smartphone data. We construct networks capturing how each user journeys across mobile Apps within a smartphone usage session, which we term ‘App-journey’ networks. Across a dataset of participants aged 18-22 (n = 82), exploratory statistical analyses show that features of ‘App-journey’ networks correlate with user mental health and attentional control. For example, users with higher depressive symptoms typically transition to a wider variety of Apps within each phone session (higher network density) and are less likely to confine each phone session to repeated groups of Apps (lower network modularity). Additionally, specific Apps which users report using more automatically are more central in App-journey networks than Apps users report using less automatically. In contrast to features of App-journey networks, time spent on one’s phone is not significantly related to mental health. In summary, this novel method to study smartphone use abstracts away from activities themselves and instead quantifies patterns of activity transitions. We show such transition patterns are related to user mental health and subjective experiences of automatic use, thus representing targets for future research and interventions.
Discretizing Pareto fronts (PFs) is essential in multi-objective optimization (MOO), enabling infinite PFs to be approximated by finite, well-distributed points. While most existing methods focus on linear PFs, the discretization of non-linear PFs remains less explored. In this paper, we introduce the Riesz s-energy-based Gradient Descent Algorithm (RSE-GDA), a novel set-based gradient descent approach to approximate near-optimal & micro;-point distributions by minimizing Riesz s-energy (Es), a diversity indicator promoting uniformity across various PF geometries. Unlike existing approaches, RSE-GDA simultaneously updates all points while preserving manifold constraints, making it effective for diverse PF geometries. Experimental results on linear, concave, and convex PFs show that RSE-GDA achieves superior uniformity and lower Es values compared to existing methods, demonstrating its effectiveness for generating uniform discretizations across diverse PF shapes.
The spatial variability in species composition, richness, and abundance of sponge recruitment within Thalassia testudinum seagrass meadows was explored using artificial seagrass units (ASUs), deployed for 85 days at the central, edge, and outer meadow zones of three sites in the southern Gulf of Mexico. Seven sponge species were recorded, matching earlier reports of adult presence in the region. Ranked in order, from highest to lowest relative abundance, species included Haliclona implexiformis , Amorphinopsis atlantica , Mycale cf. microsigmatosa , Haliclona sp., Dysidea etheria , Chondrilla caribensis , and Halichondria melanadocia . Species richness (1–3 species per ASU) and abundance (1–4 individuals per ASU) did not differ among zones within each meadow, however significant differences were found among sites, attributed to environmental differences. The detection of sponge recruits on ASUs placed at the outer zone of the meadows, where seagrass does not naturally exist, indicates that sponge larvae do disperse beyond the meadow edge, suggesting connectivity between meadows and adjacent mangrove habitats. Conversely, the absence of adult sponges in the areas outside of the meadow appears to be driven by lack of firm substrate for attachment and potentially by no seagrass canopy to shield recruits from environmental stressors, rather than by limited dispersal. These results, among the few to document sponge recruitment in seagrass habitats, contribute essential insight into larval dispersal patterns and habitat connectivity for benthic sponge fauna.
