Large Language Models Effectively Leverage Document-level Context for Literary Translation, but Critical Errors Persist
Open accessThe DOI matches its Crossref record; 22 of the 22 references it cites were found in a registry or an index.
Abstract
Large language models (LLMs) are competitive with the state of the art on a wide range of sentence-level translation datasets.However, their ability to translate paragraphs and documents remains unexplored because evaluation in these settings is costly and difficult.We show through a rigorous human evaluation that asking the GPT-3.5 (text-davinci-003) LLM to translate an entire literary paragraph (e.g., from a novel) at once results in higher-quality translations than standard sentence-by-sentence translation across 18 linguistically-diverse language pairs (e.g., translating into and out of Japanese, Polish, and English).Our evaluation, which took approximately 350 hours of effort for annotation and analysis, is conducted by hiring translators fluent in both the source and target language and asking them to provide both spanlevel error annotations as well as preference judgments of which system's translations are better.We observe that discourse-level LLM translators commit fewer mistranslations, grammar errors, and stylistic inconsistencies than sentence-level approaches.With that said, critical errors still abound, including occasional content omissions, and a human translator's intervention remains necessary to ensure that the author's voice remains intact.We publicly release our dataset and error annotations to spur future research on the evaluation of documentlevel literary translation.1
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
lmerTest Package: Tests in Linear Mixed Effects Models
Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2017) · Journal of Statistical Software · doi:10.18637/jss.v082.i13 · Cited by 24,346
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
Mixed-effects modeling with crossed random effects for subjects and items
Baayen, R. H., Davidson, D. J., & Bates, D. M. (2008) · Journal of Memory and Language · doi:10.1016/j.jml.2007.12.005 · Cited by 8,540
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988
Training language models to follow instructions with human feedback
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C. L., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. (2022) · arXiv · doi:10.48550/arxiv.2203.02155 · Cited by 4,208
BERTScore: Evaluating Text Generation with BERT
Zhang, T., Kishore, V., Wu, F., Weinberger, K. Q., & Artzi, Y. (2019) · arXiv · doi:10.48550/arxiv.1904.09675 · Cited by 2,865
GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models
Eloundou, T., Manning, S., Mishkin, P., & Rock, D. (2023) · arXiv · doi:10.48550/arxiv.2303.10130 · Cited by 2,330
Language (Technology) is Power: A Critical Survey of “Bias” in NLP
Blodgett, S. L., Barocas, S., Daumé III, H., & Wallach, H. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.485 · Cited by 1,050
BLEURT: Learning Robust Metrics for Text Generation
Sellam, T., Das, D., & Parikh, A. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.704 · Cited by 1,005
BLEURT: Learning Robust Metrics for Text Generation
Sellam, T., Das, D., & Parikh, A. P. (2020) · arXiv · doi:10.48550/arxiv.2004.04696 · Cited by 1,005
RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models
Gehman, S., Gururangan, S., Sap, M., Choi, Y., & Smith, N. A. (2020) · Findings of the Association for Computational Linguistics: EMNLP 2020 · doi:10.18653/v1/2020.findings-emnlp.301 · Cited by 693
Is ChatGPT A Good Translator? Yes With GPT-4 As The Engine
Jiao, W., Wang, W., Huang, J.-T., Wang, X., Shi, S., & Tu, Z. (2023) · arXiv · doi:10.48550/arxiv.2301.08745 · Cited by 317
Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation
Freitag, M., Foster, G., Grangier, D., Ratnakar, V., Tan, Q., & Macherey, W. (2021) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00437 · Cited by 271
How Good Are GPT Models at Machine Translation? A Comprehensive Evaluation
Hendy, A., Abdelrehim, M., Sharaf, A., Raunak, V., Gabr, M., Matsushita, H., Kim, Y. J., Afify, M., & Awadalla, H. H. (2023) · arXiv · doi:10.48550/arxiv.2302.09210 · Cited by 202
Hallucinations in Large Multilingual Translation Models
Guerreiro, N. M., Alves, D. M., Waldendorf, J., Haddow, B., Birch, A., Colombo, P., & Martins, A. F. T. (2023) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00615 · Cited by 124
Hallucinations in Large Multilingual Translation Models
Guerreiro, N. M., Alves, D., Waldendorf, J., Haddow, B., Birch, A., Colombo, P., & Martins, A. F. T. (2023) · arXiv · doi:10.48550/arxiv.2303.16104 · Cited by 124
Repairing the Cracked Foundation: A Survey of Obstacles in Evaluation Practices for Generated Text
Gehrmann, S., Clark, E., & Sellam, T. (2023) · Journal of Artificial Intelligence Research · doi:10.1613/jair.1.13715 · Cited by 123
Large Language Models Are State-of-the-Art Evaluators of Translation Quality
Kocmi, T., & Federmann, C. (2023) · arXiv · doi:10.48550/arxiv.2302.14520 · Cited by 122
The Perils of Using Mechanical Turk to Evaluate Open-Ended Text Generation
Karpinska, M., Akoury, N., & Iyyer, M. (2021) · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2021.emnlp-main.97 · Cited by 56
Self-citationDEMETR: Diagnosing Evaluation Metrics for Translation
Karpinska, M., Raj, N., Thai, K., Song, Y., Gupta, A., & Iyyer, M. (2022) · Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2022.emnlp-main.649 · Cited by 22
Self-citationBriakou, E., Cherry, C., & Foster, G. (2023) · arXiv · doi:10.48550/arxiv.2305.10266 · Cited by 19
Escaping the sentence-level paradigm in machine translation
Post, M., & Junczys-Dowmunt, M. (2023) · arXiv · doi:10.48550/arxiv.2304.12959 · Cited by 12
Cited by
A survey on large language model (LLM) security and privacy: The Good, The Bad, and The Ugly
Yao, Y., Duan, J., Xu, K., Cai, Y., Sun, Z., & Zhang, Y. (2024) · High-Confidence Computing · doi:10.1016/j.hcc.2024.100211 · Cited by 1,083
A Survey of GPT-3 Family Large Language Models Including ChatGPT and GPT-4
Kalyan, K. S. (2024) · Natural Language Processing Journal · doi:10.1016/j.nlp.2023.100048 · Cited by 335
Chen, S., & Lou, J. (2026) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2026.1727544 · Cited by 2
Metadata
- Authors
- Karpinska, M., & Iyyer, M.
- Venue
- Proceedings of the Eighth Conference on Machine Translation
- Published
- 2023
- Pages
- 419-451
- OpenAlex
- W4389519421
- Cited by
- 65
- References
- 22
- FWCI
- 12.29
- Type
- conference paper
- Language
- en
- Open access
- hybrid
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.