---
title: "LEAF: Language Learners' English Essays and Feedback Corpus"
description: "The authors build LEAF, about 6K pairs of learner essays and feedback. 41 references, cited by 2 papers here."
authors: ["Shabnam Behzad", "Omid Kashefi", "Swapna Somasundaran"]
year: 2024
venue: "Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers)"
source: "https://aclanthology.org/2024.naacl-short.36/"
doi: "10.18653/v1/2024.naacl-short.36"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/behzad-2024-leaf-language-learners-english-essays/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/behzad-2024-leaf-language-learners-english-essays/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2024](https://md.research.elamur.ai/free/en/papers/2024/page/11/) / LEAF: Language Learners' English Essays and Feedback Corpus

# LEAF: Language Learners' English Essays and Feedback Corpus

**Authors:** [Shabnam Behzad](https://md.research.elamur.ai/free/en/authors/shabnam-behzad/), Omid Kashefi and [Swapna Somasundaran](https://md.research.elamur.ai/free/en/authors/swapna-somasundaran/) · **Year:** 2024 · **Venue:** [Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers)](https://md.research.elamur.ai/free/en/venues/naacl/) ([NAACL 2024](https://md.research.elamur.ai/free/en/venues/naacl/2024/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2024.naacl-short.36/) · [PDF](https://aclanthology.org/2024.naacl-short.36.pdf) · [DOI 10.18653/v1/2024.naacl-short.36](https://doi.org/10.18653/v1/2024.naacl-short.36) · [OpenAlex](https://openalex.org/W4401042617)

**Citations:** 3 citations (OpenAlex, 2026-10-14) · [2](https://md.research.elamur.ai/free/en/papers/behzad-2024-leaf-language-learners-english-essays/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
Automated tools for learner writing mostly give overall scores or sentence-level grammar notes. Open-ended, essay-level feedback tailored to the specific text has received little attention, and no public English dataset existed for it. The authors build LEAF, about 6K pairs of learner essays and feedback. The essays come from the Writing Feedback forum of essayforum.com, which is used mostly by people preparing for English proficiency exams. Each essay has two kinds of feedback:
- the original educator comment;
- an "AI-augmented" version, in which GPT-4 rewrites its own rubric-based feedback so that it absorbs points from the human comment.

The authors then test Llama2 baselines, some with prompts holding random LEAF examples and some with retrieved examples on similar topics. Human raters judged the outputs on constructive-feedback criteria. Prompting with LEAF examples helped on nearly every criterion, and retrieving similar essays mainly improved how well the feedback identified weaknesses. Even so, the scores remain far from usable for real students. LEAF matters as a starting benchmark and training resource for personalized essay feedback.

## Key points
- **Data collection and filtering:**
  - Essays whose topic depends on an image or figure were removed, and so were overly short feedback posts.
  - Only reviewers with at least 5 feedback posts were kept.
  - Posts were also filtered by like/post ratio, with a threshold of 0.3 chosen from the distribution of ratios.
- **Basis for the AI-augmented feedback:** an earlier study by the same authors found that GPT-4 with a rubric-based prompt gave the best machine feedback, but it was overly positive. Human reviewers were more specific, more actionable, and focused on the most critical problems. This motivated merging the two sources.
- **Evaluation criteria:**
  - accuracy/relevance, added specifically to catch hallucination;
  - pointing out weaknesses;
  - pointing out strengths;
  - specificity and actionability;
  - overall helpfulness.
- **Retrieval setup:**
  - A Sentence-BERT bi-encoder (paraphrase-mpnet-base-v2) does the first search.
  - A cross-encoder (ms-marco-MiniLM-L-6-v2) reranks the results.
  - The top 3 examples go into the prompt, ordered so the closest one comes last.
- **Human evaluation and results:**
  - Setup: 20 essays × 4 baselines, giving 80 feedback texts. Six raters scored them on a 1–3 scale, with two raters per item and their scores averaged.
  - Without examples, Llama2-7B outperformed Llama2-13B. However, 13B did better on well-written essays.
  - Models often proposed unnecessary edits and praised qualities the essay did not actually have.

## Interesting ideas
- Instead of picking either human or LLM output, have the LLM revise its own draft using the human's comments, so the result keeps the LLM's broad coverage and gains the human's sharp, specific critique.
- Filter crowd-sourced labels by contributor reliability: require a minimum activity level, plus an approval-ratio threshold set from the observed distribution rather than picked arbitrarily.
- Add an explicit evaluation criterion for faithfulness to the input next to the usual quality criteria, because fluent output can confidently praise things that are absent.
- Include inputs that need little or no correction in the evaluation, since models tend to over-suggest changes there. The paper found that model rankings can even flip on such cases.
- Retrieve topically similar solved examples (fast bi-encoder search, then cross-encoder rerank, closest example last) to steer the model toward task-specific structure such as relevance to the assignment.

## For Mova

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

- **I1** (learn): an LLM teacher tier could rewrite its silver annotations after seeing disagreements from witness parsers or human hints. This would yield merged labels that are richer than either source alone.
- **I2** (learn): when pooling silver trees or external annotations, weight or drop sources by their track record. The cutoff could be set from the empirical distribution of agreement rates.
- **I3** (back): when generating text from the meaning graph, check every statement against the graph and count unsupported claims as a separate failure type.
- **I4** (learn): add "nothing to fix" items as negative controls when evaluating repair rules, so that over-eager edits are caught and penalized.
- **I5** (learn): prompts for the LLM teacher could include already-verified analyses of similar sentences or problems. This should make its proposed annotations more consistent with Mova's conventions.

*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

- [Georgetown University](https://md.research.elamur.ai/free/en/universities/georgetown-university/) — University, Washington, United States · 88 works · 4,732 citations
- [Educational Testing Service](https://md.research.elamur.ai/free/en/ngos/educational-testing-service/) — Nonprofit, Princeton, United States · 27 works · 2,212 citations

📚 Shelves (6): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works · [💭 Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/) · 1,552 works · [💭 Developmental and Educational Psychology](https://md.research.elamur.ai/free/en/shelves/psychology/developmental-and-educational-psychology/) · 499 works

🧭 Topics (2): [English](https://md.research.elamur.ai/free/en/topics/english/) · 2,263 works · [🖍️ Grammatical error correction & learner language](https://md.research.elamur.ai/free/en/topics/grammatical-error-correction/) · 156 works

## Cite

Cite as: Shabnam Behzad, Omid Kashefi and Swapna Somasundaran. 2024. LEAF: Language Learners' English Essays and Feedback Corpus. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers).

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

```bibtex
@inproceedings{behzad2024leaf,
  title = {{LEAF: Language Learners' English Essays and Feedback Corpus}},
  author = {Shabnam Behzad, Omid Kashefi and Swapna Somasundaran},
  booktitle = {Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers)},
  year = {2024},
  doi = {10.18653/v1/2024.naacl-short.36},
  url = {https://doi.org/10.18653/v1/2024.naacl-short.36}
}
```

- [← previous KSAA-CAD Shared Task: Contemporary Arabic Dictionary for Reverse Dictionary and Word Sense Disambiguation](https://md.research.elamur.ai/free/en/papers/alshammari-2024-ksaa-cad-shared-task-contemporary/)
- [next → Munazarat 1.0: A Corpus of Arabic Competitive Debates](https://md.research.elamur.ai/free/en/papers/khader-2024-munazarat-1-0-corpus-arabic/)
- [↑ back to papers of 2024](https://md.research.elamur.ai/free/en/papers/2024/page/11/)

## References

41 references, 10 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Perpetual Baffour, Tor Saxberg, and Scott Crossley. 2023. Analyzing bias in large language model solutions for assisted writing feedback tools: Lessons from the feedback prize competition series. In Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2023), pages 242–246, Toronto, Canada. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Perpetual+Baffour%2C+Tor+Saxberg%2C+and+Scott+Crossley%2E+2023%2E+Analyzing+bias+in+large+language+model+solutions+for+assisted+writing+feedback+tools%3A+Lessons+from+the+feedback+prize+competition+series%2E+In+P)
2. Lifang Bai and Guangwei Hu. 2017. In the face of fallible awe feedback: how do students respond? Educational Psychology, 37(1):67–81. [find](https://scholar.google.com/scholar?q=Lifang+Bai+and+Guangwei+Hu%2E+2017%2E+In+the+face+of+fallible+awe+feedback%3A+how+do+students+respond%3F+Educational+Psychology%2C+37%281%29%3A67%E2%80%9381%2E)
3. Jianzhu Bao, Yasheng Wang, Yitong Li, Fei Mi, and Ruifeng Xu. 2022. AEG: Argumentative essay generation via a dual-decoder model with content planning. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 5134–5148, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Jianzhu+Bao%2C+Yasheng+Wang%2C+Yitong+Li%2C+Fei+Mi%2C+and+Ruifeng+Xu%2E+2022%2E+AEG%3A+Argumentative+essay+generation+via+a+dual%2Ddecoder+model+with+content+planning%2E+In+Proceedings+of+the+2022+Conference+on+Empiric)
4. Shabnam Behzad, Omid Kashefi, and Swapna Soma-sundara. 2024. Assessing online writing feedback resources: Generative ai vs. good samaritans. In LREC-COLING. [find](https://scholar.google.com/scholar?q=Shabnam+Behzad%2C+Omid+Kashefi%2C+and+Swapna+Soma%2Dsundara%2E+2024%2E+Assessing+online+writing+feedback+resources%3A+Generative+ai+vs%2E+good+samaritans%2E+In+LREC%2DCOLING%2E)
5. Shabnam Behzad, Keisuke Sakaguchi, Nathan Schneider, and Amir Zeldes. 2023a. [ELQA: A corpus of metalinguistic questions and answers about English](https://md.research.elamur.ai/free/en/papers/behzad-2023-elqa-corpus-metalinguistic-questions-answers/). In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2031–2047, Toronto, Canada. Association for Computational Linguistics.
6. Shabnam Behzad, Amir Zeldes, and Nathan Schneider. 2023b. [Sentence-level feedback generation for English language learners: Does data augmentation help?](https://md.research.elamur.ai/free/en/papers/behzad-2023-sentence-level-feedback-generation-english/) In Proceedings of the 16th International Natural Language Generation Conference: Generation Challenges, pages 53–59, Prague, Czechia. Association for Computational Linguistics.
7. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, Chris Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Scott Gray, Benjamin Chess, Jack Clark, Christopher Berner, Sam McCandlish, Alec Radford, Ilya Sutskever, and Dario Amodei. 2020. [Language models are few-shot learners](https://md.research.elamur.ai/free/en/papers/brown-2020-language-models-few-shot-learners/). In Advances in Neural Information Processing Systems, volume 33, pages 1877–1901. Curran Associates, Inc.
8. Jill Burstein, Martin Chodorow, and Claudia Leacock. 2004. Automated essay evaluation: The Criterion online writing service. AI Magazine, 25(3):27. [↗](https://doi.org/10.1609/aimag.v25i3.1774)
9. Jill Burstein, Daniel Marcu, and Kevin Knight. 2003. Finding the write stuff: Automatic identification of discourse structure in student essays. IEEE Intelligent Systems, 18(1):32–39. [↗](https://doi.org/10.1109/mis.2003.1179191)
10. David Carless. 2006. Differing perceptions in the feedback process. Studies in higher education, 31(2):219– 233. [↗](https://doi.org/10.1080/03075070600572132)
11. Asli Celikyilmaz, Elizabeth Clark, and Jianfeng Gao. 2020. Evaluation of text generation: A survey. arXiv preprint arXiv:2006.14799. [↗](https://arxiv.org/abs/2006.14799)
12. Steven Coyne. 2023. Template-guided grammatical error feedback comment generation. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics: Student Research Workshop, pages 94–104, Dubrovnik, Croatia. Association for Computational Linguistics. [↗](https://doi.org/10.48448/wtz2-w497)
13. Jack Ende. 1983. Feedback in clinical medical education. The Journal Of The American Medical Association (JAMA), 250(6):777–781. [↗](https://doi.org/10.1001/jama.1983.03340060055026)
14. Lester Faigley and Stephen Witte. 1981. Analyzing revision. College composition and communication, 32(4):400–414. [find](https://scholar.google.com/scholar?q=Lester+Faigley+and+Stephen+Witte%2E+1981%2E+Analyzing+revision%2E+College+composition+and+communication%2C+32%284%29%3A400%E2%80%93414%2E)
15. Jiefu Gong, Xiao Hu, Wei Song, Ruiji Fu, Zhichao Sheng, Bo Zhu, Shijin Wang, and Ting Liu. 2021. IFlyEA: A Chinese essay assessment system with automated rating, review generation, and recommendation. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing: System Demonstrations, pages 240–248, Online. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Jiefu+Gong%2C+Xiao+Hu%2C+Wei+Song%2C+Ruiji+Fu%2C+Zhichao+Sheng%2C+Bo+Zhu%2C+Shijin+Wang%2C+and+Ting+Liu%2E+2021%2E+IFlyEA%3A+A+Chinese+essay+assessment+system+with+automated+rating%2C+review+generation%2C+and+recommendation%2E)
16. Jieun Han, Haneul Yoo, Junho Myung, Minsun Kim, Hyunseung Lim, Yoonsu Kim, Tak Yeon Lee, Hwajung Hong, Juho Kim, So-Yeon Ahn, et al. 2023. Fabric: Automated scoring and feedback generation for essays. arXiv preprint arXiv:2310.05191. [↗](https://arxiv.org/abs/2310.05191)
17. Wen-Bin Han, Jhih-Jie Chen, Chingyu Yang, and Jason Chang. 2019. Level-up: Learning to improve proficiency level of essays. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pages 207–212, Florence, Italy. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Wen%2DBin+Han%2C+Jhih%2DJie+Chen%2C+Chingyu+Yang%2C+and+Jason+Chang%2E+2019%2E+Level%2Dup%3A+Learning+to+improve+proficiency+level+of+essays%2E+In+Proceedings+of+the+57th+Annual+Meeting+of+the+Association+for+Computation)
18. Kazuaki Hanawa, Ryo Nagata, and Kentaro Inui. 2021. Exploring methods for generating feedback comments for writing learning. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 9719–9730, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics. [↗](https://doi.org/10.5715/jnlp.29.270)
19. John Hattie and Helen Timperley. 2007. The power of feedback. Review of educational research, 77(1):81– 112. [↗](https://doi.org/10.3102/003465430298487)
20. Omid Kashefi, Tazin Afrin, Meghan Dale, Christopher Olshefski, Amanda Godley, Diane Litman, and Rebecca Hwa. 2022. Argrewrite v. 2: an annotated argumentative revisions corpus. Language Resources and Evaluation, pages 1–35. [↗](https://doi.org/10.1007/s10579-021-09567-z)
21. Zixuan Ke and Vincent Ng. 2019. Automated essay scoring: A survey of the state of the art. In Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, IJCAI-19, pages 6300–6308. International Joint Conferences on Artificial Intelligence Organization. [find](https://scholar.google.com/scholar?q=Zixuan+Ke+and+Vincent+Ng%2E+2019%2E+Automated+essay+scoring%3A+A+survey+of+the+state+of+the+art%2E+In+Proceedings+of+the+Twenty%2DEighth+International+Joint+Conference+on+Artificial+Intelligence%2C+IJCAI%2D19%2C+page)
22. Charles W Kneupper. 1978. Teaching argument: An introduction to the toulmin model. College Composition and Communication, 29(3):237–241. [↗](https://doi.org/10.2307/356935)
23. Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, and Douwe Kiela. 2020. [Retrieval-augmented generation for knowledge-intensive nlp tasks](https://md.research.elamur.ai/free/en/papers/lewis-2020-retrieval-augmented-generation-knowledge-intensive/). In Proceedings of the 34th International Conference on Neural Information Processing Systems, NIPS’20, Red Hook, NY, USA. Curran Associates Inc.
24. Ming Liu, Yi Li, Weiwei Xu, and Li Liu. 2017. Automated essay feedback generation and its impact on revision. IEEE Transactions on Learning Technologies, 10(4):502–513. [↗](https://doi.org/10.1109/tlt.2016.2612659)
25. Pengfei Liu, Weizhe Yuan, Jinlan Fu, Zhengbao Jiang, Hiroaki Hayashi, and Graham Neubig. 2023. [Pre-train, prompt, and predict: A systematic survey of prompting methods in natural language processing](https://md.research.elamur.ai/free/en/papers/liu-2021-pre-train-prompt-predict-systematic/). ACM Comput. Surv., 55(9). [↗](https://doi.org/10.1145/3560815)
26. Grégoire Mialon, Roberto Dessì, Maria Lomeli, Christoforos Nalmpantis, Ram Pasunuru, Roberta Raileanu, Baptiste Rozière, Timo Schick, Jane Dwivedi-Yu, Asli Celikyilmaz, et al. 2023. [Augmented language models: a survey](https://md.research.elamur.ai/free/en/papers/mialon-2023-augmented-language-models-survey/). arXiv preprint arXiv:2302.07842. [↗](https://arxiv.org/abs/2302.07842)
27. Ryo Nagata. 2019. Toward a task of feedback comment generation for writing learning. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 3206–3215, Hong Kong, China. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Ryo+Nagata%2E+2019%2E+Toward+a+task+of+feedback+comment+generation+for+writing+learning%2E+In+Proceedings+of+the+2019+Conference+on+Empirical+Methods+in+Natural+Language+Processing+and+the+9th+International)
28. Ahmad Omer and Mohhamed Abdularhim. 2017. The criteria of constructive feedback: The feedback that counts. Journal of Health Specialties, 5(1):45–45. [↗](https://doi.org/10.4103/2468-6360.198798)
29. OpenAI. 2023. [Gpt-4 technical report](https://md.research.elamur.ai/free/en/papers/openai-2023-gpt-4-technical-report/).
30. Martha N. Ovando. 1994. Constructive feedback: A key to successful teaching and learning. International Journal of Educational Management, 8(6):19–22. [find](https://scholar.google.com/scholar?q=Martha+N%2E+Ovando%2E+1994%2E+Constructive+feedback%3A+A+key+to+successful+teaching+and+learning%2E+International+Journal+of+Educational+Management%2C+8%286%29%3A19%E2%80%9322%2E)
31. Ildiko Pilan, John Lee, Chak Yan Yeung, and Jonathan Webster. 2020. A dataset for investigating the impact of feedback on student revision outcome. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 332–339, Marseille, France. European Language Resources Association. [find](https://scholar.google.com/scholar?q=Ildiko+Pilan%2C+John+Lee%2C+Chak+Yan+Yeung%2C+and+Jonathan+Webster%2E+2020%2E+A+dataset+for+investigating+the+impact+of+feedback+on+student+revision+outcome%2E+In+Proceedings+of+the+Twelfth+Language+Resources+and)
32. Nils Reimers and Iryna Gurevych. 2019. [Sentence-bert: Sentence embeddings using siamese bert-networks](https://md.research.elamur.ai/free/en/papers/reimers-2019-sentence-bert-sentence-embeddings-siamese/). In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics.
33. Salvatore Valenti, Francesca Neri, and Alessandro Cucchiarelli. 2003. An overview of current research on automated essay grading. Journal of Information Technology Education: Research, 2(1):319–330. [↗](https://doi.org/10.28945/331)
34. Fan Zhang, Rebecca Hwa, Diane Litman, and Homa B. Hashemi. 2016. ArgRewrite: A web-based revision assistant for argumentative writings. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Demonstrations, pages 37–41, San Diego, California. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Fan+Zhang%2C+Rebecca+Hwa%2C+Diane+Litman%2C+and+Homa+B%2E+Hashemi%2E+2016%2E+ArgRewrite%3A+A+web%2Dbased+revision+assistant+for+argumentative+writings%2E+In+Proceedings+of+the+2016+Conference+of+the+North+American+Chap)
35. Zhe Victor Zhang and Ken Hyland. 2018. Student engagement with teacher and automated feedback on l2 writing. Assessing Writing. [↗](https://doi.org/10.1016/j.asw.2018.02.004)
36. Zhexin Zhang, Jian Guan, Guowei Xu, Yixiang Tian, and Minlie Huang. 2022. Automatic comment generation for Chinese student narrative essays. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 214–223, Abu Dhabi, UAE. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Zhexin+Zhang%2C+Jian+Guan%2C+Guowei+Xu%2C+Yixiang+Tian%2C+and+Minlie+Huang%2E+2022%2E+Automatic+comment+generation+for+Chinese+student+narrative+essays%2E+In+Proceedings+of+the+2022+Conference+on+Empirical+Methods+)
37. Nils Reimers and Iryna Gurevych. 2021. The curse of dense low-dimensional information retrieval for large index sizes. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers), pages 605–611, Online. Association for Computational Linguistics. [↗](https://doi.org/10.48448/z3am-4347)
38. Ohad Rubin, Jonathan Herzig, and Jonathan Berant. 2022. [Learning to retrieve prompts for in-context learning](https://md.research.elamur.ai/free/en/papers/rubin-2021-learning-retrieve-prompts-context-learning/). In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 2655–2671, Seattle, United States. Association for Computational Linguistics.
39. Xinshu Shen, Hongyi Wu, Xiaopeng Bai, Yuanbin Wu, Aimin Zhou, Shaoguang Mao, Tao Ge, and Yan Xia. 2023. Overview of CCL23-eval task 8: Chinese essay fluency evaluation (CEFE) task. In Proceedings of the 22nd Chinese National Conference on Computational Linguistics (Volume 3: Evaluations), pages 282–292, Harbin, China. Chinese Information Processing Society of China. [find](https://scholar.google.com/scholar?q=Xinshu+Shen%2C+Hongyi+Wu%2C+Xiaopeng+Bai%2C+Yuanbin+Wu%2C+Aimin+Zhou%2C+Shaoguang+Mao%2C+Tao+Ge%2C+and+Yan+Xia%2E+2023%2E+Overview+of+CCL23%2Deval+task+8%3A+Chinese+essay+fluency+evaluation+%28CEFE%29+task%2E+In+Proceedings+of+t)
40. Lili Tian and Yu Zhou. 2020. Learner engagement with automated feedback, peer feedback and teacher feedback in an online efl writing context. System. [↗](https://doi.org/10.1016/j.system.2020.102247)
41. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. 2023. [Llama 2: Open foundation and fine-tuned chat models](https://md.research.elamur.ai/free/en/papers/touvron-2023-llama-2-open-foundation-fine/). arXiv preprint arXiv:2307.09288. [↗](https://arxiv.org/abs/2307.09288)

*Updated 1 November 2026*