@inproceedings{hrabal-etal-2025-cuni,
    title = "{CUNI} and Phrase at {WMT}25 {MT} Evaluation Task",
    author = "Hrabal, Miroslav  and
      Glembek, Ondrej  and
      Tamchyna, Ale{\v{s}}  and
      Hildebrand, Almut Silja  and
      Eckhard, Alan  and
      {\v{S}}tola, Miroslav  and
      Penkale, Sergio  and
      {\v{S}}ime{\v{c}}kov{\'a}, Zuzana  and
      Bojar, Ond{\v{r}}ej  and
      Lavie, Alon  and
      Stewart, Craig",
    editor = "Haddow, Barry  and
      Kocmi, Tom  and
      Koehn, Philipp  and
      Monz, Christof",
    booktitle = "Proceedings of the Tenth Conference on Machine Translation",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-emnlp/2025.wmt-1.68/",
    pages = "934--944",
    ISBN = "979-8-89176-341-8",
    abstract = "This paper describes the joint effort of Phrase a.s. and Charles University{'}sInstitute of Formal and Applied Linguistics (CUNI/UFAL) on the WMT25Automated Translation Quality Evaluation Systems Shared Task. Both teamsparticipated both in a collaborative and competitive manner, i.e. they eachsubmitted a system of their own as well as a contrastive joint system ensemble.In Task{\textasciitilde}1, we show that such an ensembling{---}if chosen in a clever way{---}canlead to a performance boost. We present the analysis of various kinds ofsystems comprising both ``traditional'' NN-based approach, as well as differentflavours of LLMs{---}off-the-shelf commercial models, their fine-tuned versions,but also in-house, custom-trained alternative models. In Tasks{\textasciitilde}2 and{\textasciitilde}3 we showPhrase{'}s approach to tackling the tasks via various GPT models: Error SpanAnnotation via the complete MQM solution using non-reasoning models (includingfine-tuned versions) in Task{\textasciitilde}2, and using reasoning models in Task{\textasciitilde}3."
}