Tootfinder

Learning the Topic, Not the Language: How LLMs Classify Online Immigration Discourse Across Languages
Andrea Nasuto, Stefano Maria Iacus, Francisco Rowe, Devika Jain
https://arxiv.org/abs/2508.06435

Learning the Topic, Not the Language: How LLMs Classify Online Immigration Discourse Across Languages
Large language models (LLMs) are transforming social-science research by enabling scalable, precise analysis. Their adaptability raises the question of whether knowledge acquired through fine-tuning in a few languages can transfer to unseen languages that only appeared during pre-training. To examine this, we fine-tune lightweight LLaMA 3.2-3B models on monolingual, bilingual, or multilingual data sets to classify immigration-related tweets from X/Twitter across 13 languages, a domain character…

Multi-Orbital Charge Transfer into Nonplanar Cycloarenes Revealed with CO-Functionalized Tips
Anja Haags, Alexander Reichmann, Zilin Ruan, Qitang Fan, Larissa Egger, Hans Kirschner, Tim Naumann, Simon Werner, Olaf Kleykamp, Jose Martinez-Castro, Felix L\"upke, Fran\c{c}ois C. Bocquet, Christian Kumpf, Serguei Soubatch, Alexander Gottwald, Georg Koller, Michael G. Ramsey, Mathias Richter, J\"org Sundermeyer, Peter Puschnig, J. Michael Gottfried, F. Stefan Tautz, Sabine Wenzel

Multi-Orbital Charge Transfer into Nonplanar Cycloarenes Revealed with CO-Functionalized Tips
On-surface synthesis has allowed for the tuneable preparation of numerous molecular systems with variable properties. Recently, we demonstrated the highly selective synthesis of kekulene (>99%) on Cu(111) and isokekulene (92%) on Cu(110) from the same molecular precursor (Ruan et al., Angew. Chem. Int. Ed. 2025, e202509932). Scanning tunneling microscopy with CO-functionalized tips can identify the single molecules on the basis of their geometric structure at a low coverage on Cu(110), but it a…