---
title: "Cross-Dialect Sentence Transformation: A Comparative Analysis"
description: "Three models were prompted in two steps, a procedure the authors call \"promptuning\": a seed prompt first describes the… 11 references."
authors: ["Shruti Dutta", "Shashwat Mookherjee"]
year: 2023
venue: "arXiv:2311.07583"
source: "https://arxiv.org/abs/2311.07583"
doi: "10.48550/arxiv.2311.07583"
arxiv: "2311.07583"
licence: "CC BY-NC-SA 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by-nc-sa/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/dutta-2023-cross-dialect-sentence-transformation-comparative/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/dutta-2023-cross-dialect-sentence-transformation-comparative/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2023](https://md.research.elamur.ai/free/en/papers/2023/page/17/) / Cross-Dialect Sentence Transformation: A Comparative Analysis of Language Models for Adapting Sentences to British English

# Cross-Dialect Sentence Transformation: A Comparative Analysis of Language Models for Adapting Sentences to British English

**Authors:** Shruti Dutta and Shashwat Mookherjee · **Year:** 2023 · **Venue:** arXiv:2311.07583 · **Licence:** [CC BY-NC-SA 4.0 (arXiv)](http://creativecommons.org/licenses/by-nc-sa/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2311.07583) · [PDF](https://arxiv.org/pdf/2311.07583) · [DOI 10.48550/arxiv.2311.07583](https://doi.org/10.48550/arxiv.2311.07583) · [arXiv 2311.07583](https://arxiv.org/abs/2311.07583) · [OpenAlex](https://openalex.org/W4388717711)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
The paper tests how well large language models can rewrite sentences from American, Indian and Irish English into Standard British English. The authors built a small parallel set that pairs each dialect sentence with a British reference. The introduction says 200 sentences per dialect, but the results section says 20. Three models were prompted in two steps, a procedure the authors call "promptuning": a seed prompt first describes the source dialect and the target, and then the sentences are given one by one. The models are Llama-2-70B, Falcon-180B and Mistral, which is named both Mistral-1 and Mistral-7B-Instruct-v0.1. Each output is compared with its British reference by cosine similarity. The authors also score the untouched dialect input against the same references. They report Llama-2-70B as the best model and find Indian and Irish input closer to British English than American input. However, the unchanged input already scores 0.81, 0.93 and 0.93, and most model outputs fall below these values. The study is therefore more useful as a task framing and as a lesson in evaluation design than as proof that the models adapt dialects well.

## Key points
- Cosine scores (American / Indian / Irish): untouched input 0.81 / 0.93 / 0.93; Llama-2-70B 0.70 / 0.91 / 0.95; Falcon-180B 0.60 / 0.75 / 0.88; Mistral 0.63 / 0.93 / 0.91.
- Only Llama-2-70B on Irish (0.95) beats the do-nothing input, and Mistral only ties it on Indian; on American English every model scores clearly below the unchanged text.
- The data shows typical dialect contrasts:
  - lexical pairs such as queue/line, cinema/movies/pictures, peckish/hungry and tea/chai/tae;
  - Indian code-mixed items such as bhai, yaar and paisa vasool;
  - a sentence-final "na?" rendered as a British tag question.
- Several details are unclear or inconsistent: the dataset size (200 vs 20), the Mistral model name, which embedding the cosine similarity uses, and the Irish table, whose columns appear swapped.
- Planned future work includes more dialects, dialect-specific fine-tuning, human–AI collaboration and metrics designed for dialect translation.

## Interesting ideas
- **Identity baseline:** score the unchanged source against the reference next to every system, so you can see at once whether a transformation helps or harms; here it shows that most outputs are worse than doing nothing.
- **Parallel variant table:** one meaning written in four varieties (queue/line, pictures/movies/cinema, peckish/hungry) gives a compact map of regional lexical variation.
- **Code-mixed loanwords:** Indian English sentences mix in words like chai, yaar and bhai, and Irish spelling gives forms like tae; these are exactly the items that trip up standard pipelines.
- **Dialect discourse particles:** the Indian sentence-final "na?" is turned into a British tag question, which shows that dialects mark confirmation requests in different ways.
- **Seed prompt before items:** a fixed opening prompt states the source dialect and the target style, and the sentences follow, which keeps the instruction constant across the batch.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (learn): Every repair rule or LLM-teacher rewrite could be checked against a "leave it as is" control, and kept only if it beats that control.
- **I2** (world): The ontology could link regional variants such as queue and line to one concept, so the world of a text stays the same whatever dialect it was written in.
- **I3** (tokens): The en crate could add lemma and POS entries for common dialect loanwords and spellings, mapping chai to tea and tae to tea, so tagging and lemmas stay stable.
- **I4** (frames): The prag crate's speech-act detection could treat particles like "na?" as tag-question equivalents that ask the listener to confirm.
- **I5** (learn): LLM teacher tiers could use a fixed seed instruction per dialect or genre, so silver annotations stay consistent and teachers can be compared fairly.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Indian Institute of Technology Madras](https://md.research.elamur.ai/free/en/universities/indian-institute-of-technology-madras/) — University, Chennai, India · 16 works · 2,499 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works

🧭 Topics (2): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,894 works · [English](https://md.research.elamur.ai/free/en/topics/english/) · 2,379 works

## Cite

Cite as: Shruti Dutta and Shashwat Mookherjee. 2023. Cross-Dialect Sentence Transformation: A Comparative Analysis of Language Models for Adapting Sentences to British English. arXiv:2311.07583.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{dutta2023cross,
  title = {{Cross-Dialect Sentence Transformation: A Comparative Analysis of Language Models for Adapting Sentences to British English}},
  author = {Shruti Dutta and Shashwat Mookherjee},
  year = {2023},
  doi = {10.48550/arxiv.2311.07583},
  eprint = {2311.07583},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2311.07583}
}
```

- [← previous Content-Localization based Neural Machine Translation for Informal Dialectal Arabic: Spanish/French to Levantine/Gulf Arabic](https://md.research.elamur.ai/free/en/papers/alzamzami-2023-content-localization-based-neural-machine/)
- [next → De-Militarization of Language in Arabic Grammar Books and its Impact on Human Life](https://md.research.elamur.ai/free/en/papers/jamani-2023-de-militarization-language-arabic-grammar/)
- [↑ back to papers of 2023](https://md.research.elamur.ai/free/en/papers/2023/page/17/)

## References

11 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Google AI. Llama-2 chat. Hugging Face, 2023. [find](https://scholar.google.com/scholar?q=Google+AI%2E+Llama%2D2+chat%2E+Hugging+Face%2C+2023%2E)
2. Alexey Baevski, Markus Ott, Sergey Edunov, Michael Auli, Rafal Jozefowicz, and Veselin Stoyanov. Scaling transformer models for neural machine translation. arXiv preprint arXiv:2110.08536, 2021. [↗](https://arxiv.org/abs/2110.08536)
3. Tom B. Brown, Benjamin Mann, Niki Ryder, Marthe Subbiah, Jared Kaplan, Prafulla Dhariwal, Peter Mishkin, Ilya Chen, Maxwell Burgess, Aidan Diehl, et al. Pathways to agi: A unified approach to scaling language models. arXiv preprint arXiv:2201.08237, 2022. [↗](https://arxiv.org/abs/2201.08237)
4. Tianyi Chen, Dillon He, Jared Levine, Yuxiong Chen, Rodrigo Remedios, Ye Deng, and Yi Du. Llama: Language models for all. arXiv preprint arXiv:2302.02146, 2023. [↗](https://arxiv.org/abs/2302.02146)
5. Leo Gao, Stella Biderman, Sid Black, Laurence Golding, Travis Hoppe, Charles Foster, Jason Phang, Horace He, Anish Thite, Noa Nabeshima, Shawn Presser, and Connor Leahy. The Pile: An 800gb dataset of diverse text for language modeling. arXiv preprint, 2020. URL: https://arxiv.org/abs/2101.00027. [↗](https://arxiv.org/abs/2101.00027)
6. Technology Innovation Institute. Falcon in huggingface chat. Hugging Face, 2023. [find](https://scholar.google.com/scholar?q=Technology+Innovation+Institute%2E+Falcon+in+huggingface+chat%2E+Hugging+Face%2C+2023%2E)
7. Janna Lipenkova. Redefining conversational ai with large language models. 2023. URL: [https://towardsdatascience.com/redefining-conversational-ai-with-large-language-models-1ded1](https://towardsdatascience.com/redefining-conversational-ai-with-large-language-models-1ded1) [↗](https://towardsdatascience.com/redefining-conversational-ai-with-large-language-models-1ded1)
8. Nelson F. Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang. [Lost in the middle: How language models use long contexts](https://md.research.elamur.ai/free/en/papers/liu-2023-lost-middle-language-models-use/), 2023. URL: https://arxiv.org/abs/2307.03172. [↗](https://arxiv.org/abs/2307.03172)
9. OpenAI. Chatgpt. Hugging Face, 2023. [find](https://scholar.google.com/scholar?q=OpenAI%2E+Chatgpt%2E+Hugging+Face%2C+2023%2E)
10. Alec Radford, Sid Narang, Tim Salimans, John Schulman, Dario Amodei, Julian Clark, Karen Simonyan, Ilya Chen, Ilya Sutskever, Koray Kavukcuoglu, et al. Gpt-3.5: Scaling language models for massive code generation. arXiv preprint arXiv:2201.08237, 2022. [↗](https://arxiv.org/abs/2201.08237)
11. Facebook AI Research. Mistral in huggingface chat. Hugging Face, 2023. [7] [8] [5] [4] [3] [2] [10] [9] [1] [11] [6] [find](https://scholar.google.com/scholar?q=Facebook+AI+Research%2E+Mistral+in+huggingface+chat%2E+Hugging+Face%2C+2023%2E+%5B7%5D+%5B8%5D+%5B5%5D+%5B4%5D+%5B3%5D+%5B2%5D+%5B10%5D+%5B9%5D+%5B1%5D+%5B11%5D+%5B6%5D)

*Updated 29 October 2026*