HIPHIL Novum

HIPHIL Novum https://tidsskrift.dk/hiphilnovum HIPHIL Novum is an international, open-access, peer-reviewed, online journal for biblical linguistics. en-US Counting from volume 9 (2024), articles published in HIPHIL Novum are licensed under <a href="https://creativecommons.org/licenses/by-sa/4.0/">Attribution-ShareAlike 4.0 International (CC BY-SA 4.0)</a>. The editorial board may accept other Creative Commons licenses for individual articles, if required by funding bodies e.g. the European Research Council. With the publication of volume 9, authors retain copyright to their articles and give Hiphil Novum the right to the first publication. The authors retain copyright to earlier versions of the articles, such as the submitted and the accepted manuscript. Authors and readers may use, reuse, and build upon the published work, use it for text or data mining or for any other lawful purpose, as long as appropriate attribution is maintained. Articles in volumes 1-8 are not licensed under Creative Commons. In these volumes, all rights are reserved to the authors of the articles respectively. This implies that readers can download, read, and link to the articles, but they cannot republish the articles. Authors may post the published version of their article to their personal website, institutional repository, or a repository required by their funding agency as a part of a green open access policy. cch@dbi.edu (Christian Canu Højgaard) cch@dbi.edu (Christian Canu Højgaard) Tue, 13 Feb 2024 14:12:45 +0100 OJS 3.3.0.13 http://blogs.law.harvard.edu/tech/rss 60 Uncovering Theological and Ethical Biases in LLMs https://tidsskrift.dk/hiphilnovum/article/view/143407 This paper explores theological and ethical biases in LLMs through a novel approach involving creative text generation tasks based on biblical texts, specifically the Ten Commandments and the Book of Jonah. Utilizing models such as GPT-4 Turbo, Claude v2, PaLM 2 Chat, Llama 2 70B, and Zephyr 7B, the study employs a combination of qualitative hermeneutical analysis and quantitative textual analysis. Findings reveal a prevalent progressive bias in these models, evident in their interpretations of foundational ethical guidelines and narrative texts. This bias aligns with contemporary socio-political and environmental concerns, especially in themes of environmental ethics, social justice, and inclusivity. In the narrative task involving the Book of Jonah, a dominant interpretive trend is observed, reflecting the models' tendency to mirror historical and prevailing interpretations. This study highlights the need for multidisciplinary research into LLMs' biases, particularly their impact on religious and ethical narrative interpretation and broader societal implications. A.G. Elrod Copyright (c) 2024 A.G. Elrod https://creativecommons.org/licenses/by-sa/4.0 https://tidsskrift.dk/hiphilnovum/article/view/143407 Tue, 13 Feb 2024 00:00:00 +0100 Large Language Models and Biblical Hebrew: Limitations, pitfalls, opportunities https://tidsskrift.dk/hiphilnovum/article/view/144177 Researchers have been relying on computational methods to study Biblical Hebrew for a long time already. The recent improvements to and easy availability of Large Language Models (LLMs) like GPT prompt the question whether these models can be useful for our work as well. This paper tempers the expectations, showing that a critical analysis of earlier work exposes fundamental issues with methods involving GPT. However, depending on the task at hand a way forward with machine learning methods is possible, once we are aware of the limitations. Camil Staps Copyright (c) 2024 Camil Staps https://creativecommons.org/licenses/by-sa/4.0 https://tidsskrift.dk/hiphilnovum/article/view/144177 Tue, 04 Jun 2024 00:00:00 +0200 A novum HIPHIL Novum? https://tidsskrift.dk/hiphilnovum/article/view/143444 Christian Canu Højgaard Copyright (c) 2024 Christian Canu Højgaard https://creativecommons.org/licenses/by-sa/4.0 https://tidsskrift.dk/hiphilnovum/article/view/143444 Tue, 13 Feb 2024 00:00:00 +0100