A source-only quality model, Sentinel-src-24, predicts which texts machine translation systems will translate poorly better than heuristics, LLM judges, and expensive crowd pipelines.
In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , pages 6640–6651
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Estimating Machine Translation Difficulty
A source-only quality model, Sentinel-src-24, predicts which texts machine translation systems will translate poorly better than heuristics, LLM judges, and expensive crowd pipelines.