Abstract This article looks at the evolution of global standard setting for banks by the Basel Committee on Banking Supervision (or Basel Committee), and the challenges around the implementation of the last major package of prudential standards after the 2008 Global Financial Crisis, focusing on the last instalment of the standards called Final Basel III. While the political steer for Final Basel III was for minimal regulatory capital impact, regional and private data analyses reported much higher capital increase. The article argues that this high impact prompted concerns by industry and some policymakers, and led to local implementation with adjustments, for example in the EU and UK. The article also considers that divergent national implementations of Basel Committee standards reflect democratic accountability and local economic sensitivities. At the same time, this approach undermines the credibility of global standards and casts doubt on the future ability to undertake major regulatory reforms. The article is topical, because it looks at the case of specific banking standards against a background of broader questions about the value of global cooperation. In the case of the Basel Committee, these challenges do not come from disruptive political forces but from traditional political players who favoured local adjustments over adherence to global rules, in order to lessen the impact on their economies. The article calls for more data transparency, post-finalization feedback mechanisms, and openness to targeted adjustments to preserve global standard-setting credibility.
The increasing reliance on Large Language Models (LLMs) across diverse sectors highlights the need for robust domain-specific and language-specific evaluation datasets; however, the collection of such datasets is challenging due to privacy concerns, regulatory restrictions, and the time cost for manual creation. Existing automated benchmarking methods are often limited by relying on pre-existing data, poor scalability, single-domain focus, and lack of multilingual support. We present STELLAR-E - a fully automated system to generate high-quality synthetic datasets of custom size, using minimal human inputs without depending on existing datasets. The system is structured in two stages: (1) We modify the TGRT Self-Instruct framework to create a synthetic data engine that enables controllable, custom synthetic dataset generation, and (2) an evaluation pipeline incorporating statistical and LLM-based metrics to assess the applicability of the synthetic dataset for LLM-based application evaluations. The synthetic datasets reach an average difference of +5.7
Graph-based Retrieval-Augmented Generation (GraphRAG) extends traditional RAG by using knowledge graphs (KGs) to give large language models (LLMs) a structured, semantically coherent context, yielding more grounded answers. However, GraphRAG reasoning process remains a “black-box”, limiting our ability to understand how specific pieces of structured knowledge influence the final output. Existing explainability (XAI) methods for RAG systems, designed for text-based retrieval, are limited to interpreting an LLM’s response through the relational structures among knowledge components, creating a critical gap in transparency and trustworthiness. To address this, we introduce **XGRAG**, a novel framework that generates causally grounded explanations for GraphRAG systems by employing graph-based perturbation strategies, to quantify the contribution of individual graph components on the model’s answer. We conduct extensive experiments comparing XGRAG against RAG-Ex, an XAI baseline for standard RAG, and evaluate its robustness across various question types, narrative structures and LLMs. Our results demonstrate a 14.81\% improvement in explanation quality over the baseline RAG-Ex across NarrativeQA, FairyTaleQA, and TriviaQA, evaluated by F1-score measuring alignment between generated explanations and original answers. Furthermore, XGRAG’s explanations exhibit a strong correlation with graph centrality measures, validating its ability to capture graph structure. XGRAG provides a scalable and generalizable approach towards trustworthy AI through transparent, graph-based explanations that enhance the interpretability of RAG systems.
Abstract Financial services are a prime example of an industry subject to global standards already for years. The reasons lie in the cross-border interconnection of banks and the links between a single organization’s vulnerability with broader financial stability. The prudential framework is the set of rules that determines what and how much risk banks have, and how they can stay solvent in managing it, while performing their commercial financing activities. The framework has evolved from private and national to a detailed set of rules set at the global level, through work led by the Basel Committee on Banking Supervision (BCBS or Basel Committee). Its output has spanned the very early days of global cooperation in the 1970s to the pre-2008 global financial crisis optimism and finally to the post-crisis reforms to restore confidence and resilience in the system. Through this process, the Basel Committee has proven its willingness to internalize past experience and evolve with the times. The evolution of the Basel Committee and the development of the Basel III standards offer an example of global cooperation and standard setting for the common good of financial stability.
This cross-sectional epidemiological study aimed to provide population-based data on hypersensitivity associated with molar–incisor hypomineralisation (MIH) in 8- to 10-year-olds from Bavaria, Germany. It was hypothesized that hypersensitivity would be equally distributed among MIH teeth. A total of 5418 schoolchildren (8–10 years) were examined using the MIH criteria of the European Academy of Paediatric Dentistry (EAPD) and the MIH Treatment Need Index (MIH-TNI). MIH-TNI 1 was linked with mild MIH; MIH-TNI 2–4 corresponded to severe MIH. Hypersensitivity was recorded dichotomously (yes/no) after a two-second, 2.8-bar air blast (Schiff test). Descriptive statistics and a mixed-effects logistic regression model—adjusted for age, sex, region, tooth type, and caries status—explored hypersensitivity in MIH-affected teeth. The MIH prevalence was 17.5
