« Scientific innovation is undergoing a paradigm shift driven by the rapid advancement of Large Language Models (LLMs). As science faces mounting challenges including information overload, disciplinary silos, and diminishing returns on conventional research methods, LLMs are emerging as powerful agents capable not only of enhancing scientific workflows but also of…
« Cet été, l’EPFL et l’ETH Zurich mettront à disposition un grand modèle de langage (LLM) développé sur des infrastructures publiques. Entraîné sur le supercalculateur «Alps» du Centre Suisse de Calcul Scientifique (CSCS), ce nouveau LLM marque une étape clé pour l’IA open source et la maîtrise multilingue. (…) »
« Les outils basés sur les LLMs et l’intelligence artificielle sont devenus incontournables dans de nombreux domaines, et en particulier dans l’activité de développement informatique. Si ils peuvent être des aides efficaces, il est important d’avoir un peu de recul sur les services disponibles et les questions éthiques et juridiques associées.
« (…) In this paper, we present Q²Forge that addresses the challenge of generating new competency questions for a KG and corresponding SPARQL queries. It iteratively validates those queries with human feedback and LLM as a judge. Q²Forge is open source, generic, extensible and modular, meaning that the different modules of…
« Europe is taking control of its digital future. With the launch of LLMs4Europe, a major new project funded by the Digital Europe Programme and coordinated by ALT-EDIC, over 70 partners from across the continent have joined forces to build open, trustworthy, and multilingual Large Language Models (LLMs) tailored to European…
« Recording of the OPERAS workshop on « AI and LLMs as tools for open scholarly communication in social sciences and humanities », that took place on February 24, 2025. The workshop was co-organised by the OPERAS Special Interest Group Tools and Platforms together with the OPERAS AI Task Force. (…) »
« On December 12, 2024, the Chinese American Librarians Association(CALA) Asia Pacific Chapter hosted the fifth webinar in 2024, exploring the transformative impact of AI and large language models (LLMs) on academic search. Renowned librarian and AI enthusiast, Aaron Tay, Head of Data Services at Singapore Management University Libraries, delivered a…
« Since the release of ChatGPT in November 2022, large language models (LLMs) have seen considerable success, including in the open-source community, with many open-weight models available. However, the requirements to deploy such a service are often unknown and difficult to evaluate in advance. To facilitate this process, we conducted numerous…
« Language models can largely benefit from efficient tokenization. However, they still mostly utilize the classical BPE algorithm, a simple and reliable method. This has been shown to cause such issues as under-trained tokens and sub-optimal compression that may affect the downstream performance. We introduce Picky BPE, a modified BPE algorithm…
