Frafall fra videregående opplæring (VGO) er ikke nødvendigvis varig. Omtrent 35 prosent av alle elever som starter i VGO, slutter uten å ha fullført. 57 prosent av disse (19 prosent av alle elever) kommer tilbake til VGO innen 8 år etter fullført grunnskole. I denne artikkelen benytter vi registerdata fra Statistisk sentralbyrå for å undersøke hva som kjennetegner elever (<24 år) som kommer tilbake til VGO etter å ha sluttet, samt hvordan det går med dem senere i utdanning og arbeidsmarked. Elever som kommer tilbake, har i snitt foreldre med lavere utdanning og inntekt enn elever som ikke har opphold i løpet av VGO. De er også oftere innvandrere, har svakere karakterer fra grunnskolen og er generelt mer utenfor jobb og utdanning som voksne. Når vi sammenligner elever som kommer tilbake og fullfører VGO, med elever som kommer tilbake uten å fullføre, ser vi at tilbakevendende elever som fullfører, har flere likhetstrekk med elever som fullfører uten opphold, enn med tilbakevendende elever som ikke fullfører. De som ikke fullfører etter å ha kommet tilbake, er oftere utenfor utdanning og jobb og har lavere arbeidsinntekt som voksen enn elever som ikke kommer tilbake etter å ha sluttet.
Knowledge distillation (KD) is a core component in the training and deployment of modern generative models, particularly large language models (LLMs). While its empirical benefits are well documented -- enabling smaller student models to emulate the performance of much larger teachers -- the underlying mechanisms by which KD improves generative quality remain poorly understood. In this work, we present a minimal working explanation of KD in generative modeling. Using a controlled simulation with mixtures of Gaussians, we demonstrate that distillation induces a trade-off between precision and recall in the student model. As the teacher distribution becomes more selective, the student concentrates more probability mass on high-likelihood regions at the expense of coverage, which is a behavior modulated by a single entropy-controlling parameter. We then validate this effect in a large-scale language modeling setup using the SmolLM2 family of models. Empirical results reveal the same precision-recall dynamics observed in simulation, where precision corresponds to sample quality and recall to distributional coverage. This precision-recall trade-off in LLMs is found to be especially beneficial in scenarios where sample quality is more important than diversity, such as instruction tuning or downstream generation. Our analysis provides a simple and general explanation for the effectiveness of KD in generative modeling.
Prarthana P. Kartholy, Thandi M. Labor, Neil N. Panchal
et al.
Selecting a college major is a difficult decision for many incoming freshmen. Traditional academic advising is often hindered by long wait times, intimidating environments, and limited personalization. AI Chatbots present an opportunity to address these challenges. However, AI systems also have the potential to generate biased responses, prejudices related to race, gender, socioeconomic status, and disability. These biases risk turning away potential students and undermining reliability of AI systems. This study aims to develop a University of Maryland (UMD) A. James Clark School of Engineering Program-specific AI chatbot. Our research team analyzed and mitigated potential biases in the responses. Through testing the chatbot on diverse student queries, the responses are scored on metrics of accuracy, relevance, personalization, and bias presence. The results demonstrate that with careful prompt engineering and bias mitigation strategies, AI chatbots can provide high-quality, unbiased academic advising support, achieving mean scores of 9.76 for accuracy, 9.56 for relevance, and 9.60 for personalization with no stereotypical biases found in the sample data. However, due to the small sample size and limited timeframe, our AI model may not fully reflect the nuances of student queries in engineering academic advising. Regardless, these findings will inform best practices for building ethical AI systems in higher education, offering tools to complement traditional advising and address the inequities faced by many underrepresented and first-generation college students.
Nowadays, factors such as the ageing of the population and the prolongation of life expectancy make it mandatory to create effective social policies by considering the health and care needs of societies. This study evaluates research on long-term care insurance, which is an important determinant of social welfare, through bibliometric analysis, a quantitative research technique. On 25.08.2023, 1497 results were obtained by searching the concept of "long-term care insurance" in the Web of Science database. Because of the analysis, it was observed that the studies on long-term care insurance were distributed between 1984 and 2023. Among 112 different research fields, it was concluded that the most studies were conducted in the field of Gereontology (n=339). "Gerontologist" was the journal that published the highest number of studies (n=76) on long-term care. The National Centre for Geriatrics Gerontology published the most studies (n=103); Ichiro Tsuji is the author of the most studies with 53 publications. The report "2018 Alzheimer’s disease facts and figures" is the most cited study. The study titled “Care-Needs Certification in Long-Term Care Insurance System of Japan”, carried out by Takako Tsutsui and Naoko Muramatsu (2005) independently of institutions, has the most citations. "Long-term care insurance" (n=332), "long-term care" (n=225) and "Japan" (n=104) are among the most frequently used keywords.
Industrial relations, Social insurance. Social security. Pension
The Working Formal Methods Symposium (FROM) is a series of workshops that aim to bring together researchers and practitioners who work on formal methods by contributing new theoretical results, methods, techniques, and frameworks, and/or by creating or using software tools that apply theoretical contributions.
Sofia Malmrud, Helena Falkenberg, Petra Lindfors
et al.
This study explored pay-setting managers’ experiences regarding the individualized pay-setting process. Seven semi-structured group-interviews with pay-setting managers (N = 28) from four private companies in Sweden were conducted. A thematic analysis identified three main themes: 1) Prerequisites for pay-setting, which included conditions for pay-setting work and experiences of these conditions; 2) Assessment and feedback, which included experiences of employee performance assessment and feedback provision; 3) Rewards, which covered experiences of different pay incentives and the relationship between performance and pay. The pay-setting process was considered to include many obstacles as well as a few opportunities. Without proper pre-requisites to assess employee performance, the possibilities to adequately reward performance were experienced as limited, which, in turn, hampered possibilities to justify both the assessment and pay raise. Taken together, this study underscores the conflict between intentions relating to how to carry out a pay-setting process and managers’ difficulties to actually accomplish this.
We use administrative panel data on the universe of Brazilian formal workers to investigate the labor market effects of the Venezuelan crisis in Brazil, focusing on the border state of Roraima. The results using difference-in-differences show that the monthly wages of Brazilians in Roraima increased by around 2 percent, which was mostly driven by those working in sectors and occupations with no refugee involvement. The study finds negligible job displacement for Brazilians but finds evidence of native workers moving to occupations without immigrants. We also find that immigrants in the informal market offset the substitution effects in the formal market.
We introduce Stream-K, a work-centric parallelization of matrix multiplication (GEMM) and related computations in dense linear algebra. Whereas contemporary decompositions are primarily tile-based, our method operates by partitioning an even share of the aggregate inner loop iterations among physical processing elements. This provides a near-perfect utilization of computing resources, regardless of how efficiently the output tiling for any given problem quantizes across the underlying processing elements. On GPU processors, our Stream-K parallelization of GEMM produces a peak speedup of up to 14$\times$ and 6.7$\times$, and an average performance response that is both higher and more consistent across 32,824 GEMM problem geometries than state-of-the-art math libraries such as CUTLASS and cuBLAS. Furthermore, we achieve this performance from a single tile size configuration per floating-point precision, whereas today's math libraries employ complex kernel-selection heuristics to select from a large ensemble of kernel variants.
Artificial Intelligence (AI), and in particular generative models, are transformative tools for knowledge work. They problematise notions of creativity, originality, plagiarism, the attribution of credit, and copyright ownership. Critics of generative models emphasise the reliance on large amounts of training data, and view the output of these models as no more than randomised plagiarism, remix, or collage of the source data. On these grounds, many have argued for stronger regulations on the deployment, use, and attribution of the output of these models. However, these issues are not new or unique to artificial intelligence. In this position paper, using examples from literary criticism, the history of art, and copyright law, I show how creativity and originality resist definition as a notatable or information-theoretic property of an object, and instead can be seen as the property of a process, an author, or a viewer. Further alternative views hold that all creative work is essentially reuse (mostly without attribution), or that randomness itself can be creative. I suggest that creativity is ultimately defined by communities of creators and receivers, and the deemed sources of creativity in a workflow often depend on which parts of the workflow can be automated. Using examples from recent studies of AI in creative knowledge work, I suggest that AI shifts knowledge work from material production to critical integration. This position paper aims to begin a conversation around a more nuanced approach to the problems of creativity and credit assignment for generative models, one which more fully recognises the importance of the creative and curatorial voice of the users of these models and moves away from simpler notational or information-theoretic views.
Anastasiia Tkalich, Nils Brede Moe, Nina Haugland Andersen
et al.
Pair programming (PP) has been a widespread practice for decades and is known for facilitating knowledge exchange and improving the quality of software. Many agilists advocated the importance of collocation, face-to-face interaction, and physical artifacts incorporated in the shared workspace when pairing. After a long period of forced work-from-home, many knowledge workers prefer to work remotely two or three days per week, which is affecting practices such as PP. In this revelatory single-case study, we aimed to understand how PP is practiced during hybrid work when team members alternate between on-site days and working from home. We collected qualitative and quantitative data through 11 semi-structured interviews, observations, feedback sessions, and self-reported surveys. The interviewees were members of an agile software development team in a Norwegian fintech company. The results presented in this paper indicate that PP can be practiced through on-site, remote, and mixed sessions, where the mixed mode seems to be the least advantageous. The findings highlight the importance of adapting the work environment to suit individual work mode preferences when it comes to PP. In the future, we will build on these findings to explore PP in other teams and organizations practicing hybrid work.
Abstract Walter Rodney is a household name in IPE. Traditionally, scholars place Rodney in the tradition of dependency theory. More recently, scholars of ‘racial capitalism’ have identified Rodney as a foundational analyst of the empirical and conceptual relationship of racism and capitalism. However, while Rodney’s work is often invoked in contemporary literature, it is rarely discussed. This paper offers a detailed engagement with Rodney’s major works and situates them in their broader intellectual context in the history of IPE. In doing so, the paper highlights three major contributions. First, Rodney locates processes of working-class formation within the historical emergence and structural constraints of the capitalist world economy, breaking down conventional divides between domestic and international political economy. Second, Rodney demonstrates the centrality of unfree labor to the extension and reproduction of capitalism on a world scale. Third, Rodney highlights the integral role of race and racism in working-class formation, but stops short of positing these relations as functionally necessary to capitalist reproduction – a nuance which distinguishes him from current fashions. The paper argues that retrieving these three contributions can help renew a critical-historical IPE centered on the development of – and challenges to – working-class power in the global economy.
Purpose: This research tries to reach an evaluation model for sustainable performance of employees in Mostazafeen Foundation Properties and Lands Organization.
Methodology: The research method was both quantitative and qualitative. The statistical population in the qualitative section included texts and themes, as well as organizational and academic experts. By applying snowball method as a sampling technique for in-depth interviews, seven organizational experts and seven academic experts were selected. The statistical population in the quantitative part included all employees and managers of this organization, from among of whom 152 individuals were selected as the statistical sample using Krejcie and Morgan table. Also, using Fuzzy Delphi technique, the statistical sample in the validation section of the identified components consisted of 10 academic experts who were selected by purposive judgment method. Then by using confirmatory factor analysis, prioritization of dimensions, components and indicators of the model were tested.
Findings: The proposed model contains three dimensions, eight components and twenty one indicators, which can be used as a proper framework for evaluating the sustainable performance of human resources. Besides, the model shows that the highest to the lowest ranks among the dimensions are respectively job quality, success, and sustainable behavior.
Originality: The presented model can help develop the existing literature and to identify the characteristics of employees' sustainable performance.
Implications: Testing the model showed that among the three main dimensions, sustainable behaviors including citizenship behavior and green behavior of employees have the lowest ranking. Therefore, it is suggested that the managers of the organization make plans to solve this problem.
Economic growth, development, planning, Employee participation in management. Employee ownership. Industrial democracy. Works councils
Maxwell's demons work by rectifying thermal fluctuations. They are not expected to function at macroscopic scales where fluctuations become negligible and dynamics become deterministic. We propose an electronic implementation of an autonomous Maxwell's demon that indeed stops working in the regular macroscopic limit as the dynamics becomes deterministic. However, we find that if the power supplied to the demon is scaled up appropriately, the deterministic limit is avoided and the demon continues to work. The price to pay is a decreasing thermodynamic efficiency. Our work suggests that novel strategies may be found in nonequilibrium settings to bring to the macroscale non-trivial effects so far only observed at microscopic scales.
Identifying the factors that influence labor force participation could elucidate how individuals arrive at their labor supply decisions, whose understanding is, in turn, of crucial importance in analyzing how the supply side of the labor market functions. This paper investigates the effect of parenthood status on Labor Force Participation (LFP) decisions using an individual-level fixed-effects identification strategy. The differences across individuals and over time in having or not having children as well as being or not being in the labor force provide the variation needed to assess the association between individuals' LFP behavior and parenthood. Parenthood could have different impacts on mothers than it would on fathers. In order to look at the causal effect of maternity and paternity on LFP separately, the data is disaggregated by gender. To this end, the effect of a change in the parenthood status can be measured using individual-level fixed-effects to account for time-invariant characteristics of individuals becoming a parent. The primary data source used is the National Longitudinal Surveys (NLS). Considering the nature of LFP variable, this paper employs Binary Response Models (BRMs) to estimate LFP equations using individual-level micro data. The findings of the study show that parenthood has a negative overall effect on LFP. However, paternity has a significant positive effect on the likelihood of being in the labor force, whilst maternity has a significant negative impact of LFP. In addition, the results imply that the effect of parenthood on LFP has been fading away over time, regardless of the gender of parents. These two pieces of evidence precisely map onto the theoretical predictions made by the related mainstream economic theories (the traditional neoclassical theory of labor supply as well as Becker's household production model). These results are ...
As more individuals consider permanently working from home, the online labor market continues to grow as an alternative working environment. While the flexibility and autonomy of these online gigs attracts many workers, success depends critically upon self-management and workers' efficient allocation of scarce resources. To achieve this, freelancers may develop alternative work strategies, employing highly standardized schedules and communication patterns while taking on large work volumes, or engaging in smaller numbers of jobs whilst tailoring their activities to build relationships with individual employers. In this study, we consider this contrast in relation to worker communication patterns. We demonstrate the heterogeneous effects of standardization versus personalization across different stages of a project and examine the relative impact on job acquisition, project completion, and earnings. Our findings can inform the design of platforms and various worker support tools for the gig economy.
AbstractThis paper focuses on the ideology and discourses of Tekstil İṣçileri Sendikası (the Textile Workers’ Union, Tekstil) in Turkey to highlight some of the specific visions of the organized labor for an emancipatory gender politics during the 1970s. This history of intersection between gender and working-class organizing has been overlooked by the Left scholarship on the one hand and liberal feminist scholarship on the other. This paper addresses this gap in the literature by highlighting gender and class concurrently throughout the history of the transformation of gender politics in labor organizations. The history of the simultaneous development of gender-related policies in Tekstil/DİSK and TEKSİF/Türk-İṣ reveals an unexplored aspect of the contentious dynamic between rival labor organizations. Between 1975–1980, the politics of gender became another pillar in trade union competition. Following the transnational influences in this transformation, this paper highlights a forgotten period of labor organizing and locates it within the history of labor and women's movements at the national and global scale.
This dossier gathers eight articles concerned with company welfare plans in Argentina, Chile, and Brazil in the twentieth century. Based on well-researched case studies, they point to the relevance of these programs in the Southern Cone and call attention to their mutable nature. The articles delve into the complex relationship between capital and labor to assess how class conflicts and class cooperation intertwined in the vast diversity of Latin American labor worlds. Although the historical analysis of these studies circumscribes at a local or a national level, the dossier looks for encouraging dialogues and exchanges among scholars interested in comparing and contrasting different national and industrial experiences in this region. The articles remind us of the importance of thinking about industrial paternalism from a historical perspective to recover the trial and error path companies pursued on designing welfare policies toward workers and employees. In the contested process of making these corporative programs, these studies highlight the agency of employers, experts, state officials, white-collar employees, and workers. Lastly, they explore the scope and limitations of these policies in changing economic, social, political, and even cultural conjunctures. In so doing, the dossier invites to rethink how social scientists, particularly historians, conceptualize these policies of extra-salary compensations by revisiting the categories of industrial paternalism, patronage, and company welfare, among others. For its organization, this dossier follows a double criterion. It combines studies that deal with analogous economic sectors in similar historical contexts. At the same time, it underscores the value of assessing changes and continuities in terms of these companies’ programs in the unpredictable and turbulent twentieth century.