In recent years, cybersecurity threats have increasingly exploited human behaviour rather than purely technical vulnerabilities, exposing the limits of traditional awareness programmes delivered outside real-world contexts. To bridge this gap, we introduce TrainShield, an interaction paradigm for contextual cybersecuri...
Giovanni Pizzenti, Alberto Verna, Nikhil Jha et al.· Proceedings of the 37th ACM...· 0 citations
Reproduction materials for a study of Legal Judgment Prediction (LJP) on Korean drunk-driving cases: an attorney-verified benchmark of 1,500 cases drawn from the LBox Open corpus, eight classical machine-learning configurations, four large-language-model configurations, and a comparison of the two model families on bot...
Sungsoo Choi, Taeyeon Oh· Zenodo (CERN European Organi...· 0 citations
High quality data on natural hazard damages are crucial for effective disaster risk management. Yet, existing impact datasets remain limited and often biased toward Northern countries and monetary losses. To help address these gaps, we present ROUGE (Redcross Operations Unified Global Emergency database); a new socio...
L. Hasbini, L. Severino, M. M. de Brito et al.· Scientific Data· 0 citations
This paper develops a dynamic framework for analyzing civilizational reorganization by extending the theory of reaction synchronization along a temporal axis. Rather than treating civilizations as unified entities that rise, decline, and replace one another, it conceptualizes civilization as an overlapping assemblage o...
KOUHEI MISUGI· Zenodo (CERN European Organi...· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
EuLLM is an open-source platform for creating, distributing, and running sovereign EU-compliant Large Language Models, designed for verticalization across domains, languages, and brands while ensuring AI Act compliance. The platform consists of three components: Engine — a Rust-based inference runtime built on top of l...
Francesco Marchetti· Zenodo (CERN European Organi...· 0 citations
Position paper proposing a coherence-monitoring approach to adversarial attacks on large language models, with an explicit falsification condition. No experimental results are reported. v1.2 corrects metadata and citation errors from v1.0, tightens the scope of the compression-attack signature, operationalizes the cont...
Cody A Kristenson· Zenodo (CERN European Organi...· 0 citations
Second edition, September 2026. This version corrects substantive errors in thefirst edition (17 January 2026) and deposits the surviving primary recordsalongside the manuscript. A list of changes appears in the "Note on thisedition" section of the paper. Background: The introduction of large language models (LLMs) int...
Taras Shlyakhta· Zenodo (CERN European Organi...· 0 citations
With the advent of large multimodal language models, science is now at a threshold of an AI-based technological transformation. An emerging ecosystem of models and tools aims to support researchers throughout the scientific lifecycle, including (1) searching for relevant literature, (2) generating research ideas and co...
Steffen Eger, Cao Yong, Jennifer D’Souza et al.· ACM Computing Surveys· 7 citations
Abstract: Aligned large language models demonstrate epistemic overriding, flattening human ambiguity into assertive, unprompted premises. This behavior is the structural inverse of sycophancy. It occurs because confident disagreement in a language model is an engineered format rather than internal reasoning. Alignment...
Josie Jefferson, Felix Velasco· Zenodo (CERN European Organi...· 0 citations
Abstract: Aligned large language models demonstrate epistemic overriding, flattening human ambiguity into assertive, unprompted premises. This behavior is the structural inverse of sycophancy. It occurs because confident disagreement in a language model is an engineered format rather than internal reasoning. Alignment...
Josie Jefferson, Felix Velasco· Zenodo (CERN European Organi...· 0 citations
Large language models are billed by the token, and tokens are not distributed evenly across languages. We measure how many tokens ten widely deployed tokenizers require to encode identical content in Kazakh and English, using 900 sentence-aligned pairs from OPUS-100 with bootstrap confidence intervals over 10,000 resam...
Alikhan Akim, Zhibek Sarsen· Zenodo (CERN European Organi...· 0 citations
