SLoW: Select Low-frequency Words! Automatic Dictionary Selection for Translation on Large Language Models
Abstract
There are more than 7,000 languages around the world, and current Large Language Models (LLMs) only support hundreds of languages. Dictionary-based prompting methods can enhance translation on them, but most methods use all the available dictionaries, which could be expensive. Instead, it will be flexible to have a trade-off between token consumption and translation performance. This paper proposes a novel task called Automatic Dictionary Selection (ADS). The goal of the task is to automatically select which dictionary to use to enhance translation. We propose a novel and effective method which we call Select Lowfrequency Words! (SLoW) which selects those dictionaries that have a lower frequency. Our methods have unique advantages. First, there is no need for access to the training data for frequency estimation (which is usually unavailable). Second, it inherits the advantage of dictionary-based methods, where no additional tuning is required on LLMs. Experimental results on 100 languages from FLORES indicate that SLoW surpasses strong baselines, and it can obviously save token usage, with many languages even surpassing the translation performance of the full dictionary baseline. 12 Ground-truth Tsonga Muteki wa swifaniso u tleketliwile ku ya le Ronald Reagan UCLA Medical Center, laha a fikeke a fa. Ground-truth English The photographer was transported to Ronald Reagan UCLA Medical Center, where he subsequently died. Vanilla Model COMET Score: 0.579 ✗ "Ambulance from Ronald Reagan UCLA Medical Center has arrived." Differ in Round-trip COMET Score: 0.413 ✗ The ensued entertainment after Ronald Reagan UCLA Medical Centre called the police but attempted him for throad and for traffic stop due to him driving too fast therefore prompting visual imagery for the first time in South Africa. Differ in Translation COMET Score: 0.335 ✗ The driver of the police vehicle had his ear given according to the witness who is accused of not being prepared to pay a fine. Noun Dictionary COMET Score: 0.806 ✗ The person in the photographs was taken to the Ronald Reagan UCLA Medical Center, where he died. Adjective Dictionary COMET Score: 0.562 ✗ Ronald Reagan Center where he died. Noun and Adjective Dictionary COMET Score: 0.763 ✗ The driver of the vehicle was taken to Ronald Reagan UCLA Medical Center, where he died. Noun, Adjective, and Verb Dictionary COMET Score: 0.864 ✗ The photographer was taken to the Ronald Reagan UCLA Medical Center, where he died. High-frequency COMET Score: 0.707 ✗ The man of images was taken to the Ronald Reagan UCLA Medical Center, where he died. SLoW COMET Score: 0.912 ✓ The photographer was transported to the Ronald Reagan UCLA Medical Center, where he subsequently died.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
1 top-tier paper has cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Large Language Models Can Be Easily Distracted by Irrelevant Context cites Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters
- Large Language Models Can Be Easily Distracted by Irrelevant Context cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- Prompting PaLM for Translation: Assessing Strategies and Performance cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- Adam's Law: Textual Frequency Law on Large Language Models cites Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters
- Adam's Law: Textual Frequency Law on Large Language Models cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
Cited by top-tier papers1
Builds on10
- Chain-of-Thought Prompting Elicits Reasoning in Large Language ModelsJason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma et al.NeurIPS 2022 · 22,562 citations
- Large Language Models Can Be Easily Distracted by Irrelevant ContextFreda Shi, Xinyun Chen, Kanishka Misra, Nathan Scales et al.ICML 2023 · 970 citations
- Few-shot Learning with Multilingual Generative Language ModelsXi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang et al.EMNLP 2022 · 113 citations
- Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What MattersBoshi Wang, Sewon Min, Xiang Deng, Jiaming Shen et al.ACL 2023 · 100 citations
- Prompting PaLM for Translation: Assessing Strategies and PerformanceDavid Vilar, Markus Freitag, Colin Cherry, Jiaming Luo et al.ACL 2023 · 70 citations
Related papers
- Compositional Translation: A Novel LLM-based Approach for Low-resource Machine TranslationArmel Randy Zebaze, Benoît Sagot, Rachel BawdenEMNLP Findings 2025 · 4 citations
- Chain-of-Dictionary Prompting Elicits Translation in Large Language ModelsHongyuan Lu, Haoran Yang, Haoyang Huang, Dongdong Zhang et al.EMNLP 2024 · 9 citations
- No One Fits All: From Fixed Prompting to Learned Routing in Multilingual LLMsWei-Chi Wu, Sheng-Lun Wei, Hen-Hsen Huang, Hsin-Hsi ChenACL Findings 2026
- GrammaMT: Improving Machine Translation with Grammar-Informed In-Context LearningRita Ramos, Everlyn Asiko Chimoto, Maartje ter Hoeve, Natalie SchluterACL 2025 · 10 citations
- Adaptive Few-shot Prompting for Machine Translation with Pre-trained Language ModelsLei Tang, Jinghui Qin, Wenxuan Ye, Hao Tan et al.AAAI 2025 · 9 citations
