A collection of my research papers, preprints and publications. Each entry links to the paper, code and other materials where available.
LocQE: Principled Domain Adaptation for Localisation Quality Estimation by Leveraging Post-Edits
Kathy Hämmerl, Gabriel Bretschner, Joern WuebkerSep 16, 2026arXiv preprint arXiv:2609.18720
Adapting quality estimation to real-world localisation by fine-tuning on post-edits, so models better rank preferred translations of the same source.
quality estimationmachine translationdomain adaptation
Terminal-Bench-LILT: Multilingual Agentic Coding Benchmark Grounded in Language, Region, and Culture
Yunsu Kim, Kaden Uhlig, Ashwin Purohit, Milind Agarwal, Patrick Simianer, Anil Arslan, Kiarash Mokhtari, Thomas Zenkel, Johannes Mosig, Gabriel Bretschner, Shamik Bose, Joern Wuebker, John DeNeroAug 11, 2026arXiv preprint arXiv:2608.28641
A benchmark of 300 authentic coding tasks across ten languages, showing that multilingual coding competence is a distinct capability for modern coding agents.
multilingualcoding agentsbenchmark
On the alignment problem in multi-head attention-based neural machine translation
Tamer Alkhouli, Gabriel Bretschner, Hermann NeyOct 1, 2018Proceedings of the Third Conference on Machine Translation: Research Papers (WMT 2018)
Adding a dedicated alignment head to transformer attention for sharper alignments, dictionary-guided translation and faster decoding.
neural machine translationattentionword alignment
Alignment-based neural machine translation
Tamer Alkhouli, Gabriel Bretschner, Jan-Thorsten Peter, Mohammed Hethnawi, Andreas Guta, Hermann NeyAug 1, 2016Proceedings of the First Conference on Machine Translation: Volume 1, Research Papers (WMT 2016)
Bridging conventional word alignment models and neural machine translation: a neural alignment model combined with a lexical model in a word-based decoder.
neural machine translationword alignmentHMM