{"ID":2878786,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2508.18549","arxiv_id":"2508.18549","title":"COMET-poly: Machine Translation Metric Grounded in Other Candidates","abstract":"Automated metrics for machine translation attempt to replicate human judgment. Unlike humans, who often assess a translation in the context of multiple alternatives, these metrics typically consider only the source sentence and a single translation. This discrepancy in the evaluation setup may negatively impact the performance of automated metrics. We propose two automated metrics that incorporate additional information beyond the single translation. COMET-polycand uses alternative translations of the same source sentence to compare and contrast with the translation at hand, thereby providing a more informed assessment of its quality. COMET-polyic, inspired by retrieval-based in-context learning, takes in translations of similar source texts along with their human-labeled quality scores to guide the evaluation. We find that including a single additional translation in COMET-polycand improves the segment-level metric performance (0.079 to 0.118 Kendall's tau-b correlation), with further gains when more translations are added. Incorporating retrieved examples in COMET-polyic yields similar improvements (0.079 to 0.116 Kendall's tau-b correlation). We release our models publicly.","short_abstract":"Automated metrics for machine translation attempt to replicate human judgment. Unlike humans, who often assess a translation in the context of multiple alternatives, these metrics typically consider only the source sentence and a single translation. This discrepancy in the evaluation setup may negatively impact the per...","url_abs":"https://arxiv.org/abs/2508.18549","url_pdf":"https://arxiv.org/pdf/2508.18549v1","authors":"[\"Maike Züfle\",\"Vilém Zouhar\",\"Tu Anh Dinh\",\"Felipe Maia Polo\",\"Jan Niehues\",\"Mrinmaya Sachan\"]","published":"2025-08-25T22:55:22Z","proceeding":"cs.CL","tasks":"[\"cs.CL\"]","methods":"[]","has_code":false}