CompactQE: Interpretable Translation Quality Estimation via Small Open-Weight LLMs
The paper titled 'CompactQE' presents a new approach to translation quality estimation using smaller, open-source language models. These models are shown to be effective and cost-efficient alternatives to larger proprietary models, addressing data privacy concerns. The authors demonstrate that their models achieve competitive results in quality assessment compared to traditional methods and human evaluations.
- ▪Current translation quality estimation relies heavily on large, proprietary language models, which raise privacy issues.
- ▪The authors propose using smaller, open-source language models with less than 30 billion parameters as a viable alternative.
- ▪Their models can generate quality scores, error annotations, and suggested corrections in a single pass, outperforming traditional metrics.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Computation and Language arXiv:2605.15763 (cs) [Submitted on 15 May 2026] Title:CompactQE: Interpretable Translation Quality Estimation via Small Open-Weight LLMs Authors:Kamil Guttmann, Zofia Fraś, Artur Nowakowski, Krzysztof Jassem View a PDF of the paper titled CompactQE: Interpretable Translation Quality Estimation via Small Open-Weight LLMs, by Kamil Guttmann and 3 other authors View PDF HTML (experimental) Abstract:Current state-of-the-art Quality Estimation (QE) in machine translation relies on massive, proprietary LLMs, raising data privacy concerns. We demonstrate that smaller, open-source LLMs (<30B parameters) are a viable, cost-effective and privacy-preserving alternative.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
