Iskanje po repozitoriju

Opis: Large language models (LLMs) have become central to both academic research and industrial applications, fueling debates on their accuracy, usability, privacy, and potential misuse. While proprietary models benefit from substantial investments in data and computing resources, open-sourcing is often suggested as a means to enhance trust and transparency. Yet, open-sourcing comes with its own challenges, such as risks of illicit applications, limited financial incentives, and intellectual property concerns. Positioned between these extremes are hybrid approaches—including partially open models and licensing restrictions—that aim to balance openness with control. In this paper, we adopt a data-driven approach to examine the open-source development of LLMs. By analyzing contributions in model improvements, modifications, and methodologies, we assess how community efforts impact model performance. Our findings indicate that the open-source community can significantly enhance models, demonstrating that community-driven modifications can yield efficiency gains without compromising performance. Moreover, our analysis reveals distinct trends in community growth and highlights which architectures benefit disproportionately from open-source engagement. These insights provide an empirical foundation to inform balanced discussions among industry experts and policymakers on the future direction of AI development.
Ključne besede: large language models, artificial intelligence, open source, data science, HuggingFace
Objavljeno v RUP: 25.09.2025; Ogledov: 600; Prenosov: 4
Celotno besedilo (606,12 KB)
Gradivo ima več datotek! Več...

Opis: Life-Cycle Assessment (LCA) is a methodology that is used extensively for evaluating the environmental impacts of products and processes throughout their lifetime. The method is highly dependent on the quality and accuracy of the underlying data. Moreover, the data acquisition process can be subjective, raising concerns about potential inconsistencies. In this study, we perform Benford’s law conformity tests (first digit) on all numerical data in ecoinvent, focusing on individual compartments (air, water, soil, and natural resources) and environmental elementary flows (carbon, toxic substances, greenhouse gases, and heavy metals), and discrepancies across continents are examined. Life Cycle Inventory data met the requirements of Benford’s law and generally exhibited high conformity. Substantial differences in conformity were observed between Africa and Europe. Individual processes and measurements were inspected to further isolate potential sources of the non-conformity. The statistical significance of the results was increased using open-source databases available on OpenLCA Nexus, including WorldSteel, OzLCI2019, ELCD, NEEDS, BioenergieDat, and Exiobase. Finally, the Environmental Performance Index (EPI) was used, and a strong correlation between continental Benford conformity results and corresponding EPI scores was observed. The findings suggest that discrepancies in conformity across continents reflect differences in data transparency and reporting practices. European datasets generally show higher conformity, likely owing to the use of more standardized methodologies. In contrast, data from regions with limited infrastructure or less established LCA practices tend to show lower conformity. Benford’s Law offers a simple and computationally efficient alternative to conventional data quality assessments without requiring additional metadata or probabilistic modeling. Its application can support the detection of systemic biases and improve the reliability of LCA-based indicators such as environmental product declarations.
Ključne besede: anomaly detection, Benford’s law, data integrity
Objavljeno v RUP: 12.06.2025; Ogledov: 2335; Prenosov: 35
Celotno besedilo (1,41 MB)
Gradivo ima več datotek! Več...