CompactQE: Interpretable Translation Quality Estimation via Small Open-Weight LLMs
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917499729084416 |
|---|---|
| author | Guttmann, Kamil Fraś, Zofia Nowakowski, Artur Jassem, Krzysztof |
| author_facet | Guttmann, Kamil Fraś, Zofia Nowakowski, Artur Jassem, Krzysztof |
| contents | 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. Using a single-pass prompting strategy, our models simultaneously generate quality scores, MQM error annotations, suggested error corrections, and full post-editions. Our analysis shows these models achieve highly competitive system-level correlations with human judgments that outperform traditional neural metrics, fine-tuned models, and human inter-annotator agreement, effectively approximating the capabilities of much larger proprietary LLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_15763 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CompactQE: Interpretable Translation Quality Estimation via Small Open-Weight LLMs Guttmann, Kamil Fraś, Zofia Nowakowski, Artur Jassem, Krzysztof Computation and Language Artificial Intelligence 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. Using a single-pass prompting strategy, our models simultaneously generate quality scores, MQM error annotations, suggested error corrections, and full post-editions. Our analysis shows these models achieve highly competitive system-level correlations with human judgments that outperform traditional neural metrics, fine-tuned models, and human inter-annotator agreement, effectively approximating the capabilities of much larger proprietary LLMs. |
| title | CompactQE: Interpretable Translation Quality Estimation via Small Open-Weight LLMs |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2605.15763 |