---
title: "基于大语言模型的中文实体链接实证研究 — 徐正斐 2025"
description: "The study asks whether the emergent abilities of large language models (step-by-step reasoning, instruction following, in-context learning) can improve…"
authors: ["徐正斐", "辛欣"]
year: 2025
venue: "自动化学报 (Acta Automatica Sinica)"
source: "https://www.aas.net.cn/cn/article/doi/10.16383/j.aas.c240069"
doi: "10.16383/j.aas.c240069"
licence: "Free access (aas.net.cn), no open licence stated — readonly; © 《自动化学报》编辑部"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/papers/xu-2025-empirical-study-chinese/"
---
[Elamur Research](https://md.research.elamur.ai/) · [this page in HTML](https://research.elamur.ai/free/papers/xu-2025-empirical-study-chinese/)

[home](https://md.research.elamur.ai/) / [papers](https://md.research.elamur.ai/free/papers/) / [2025](https://md.research.elamur.ai/free/papers/2025/page/3/) / 基于大语言模型的中文实体链接实证研究

# 基于大语言模型的中文实体链接实证研究

**Authors:** 徐正斐, 辛欣 · **Year:** 2025 · **Venue:** 自动化学报 (Acta Automatica Sinica) · **Licence:** Free access (aas.net.cn), no open licence stated — readonly; © 《自动化学报》编辑部

**Links:** [Publisher's page](https://doi.org/10.16383/j.aas.c240069) · [PDF](https://www.aas.net.cn/cn/article/pdf/preview/10.16383/j.aas.c240069.pdf) · [DOI 10.16383/j.aas.c240069](https://doi.org/10.16383/j.aas.c240069)

## Summary
The study asks whether the emergent abilities of large language models (step-by-step reasoning, instruction following, in-context learning) can improve Chinese entity linking, where encoder models such as BERT tend to overfit and handle rare entities poorly. The authors build a two-stage retrieve-and-rerank pipeline. A dual encoder proposes candidate entities through maximum inner product search, and an LLM then disambiguates among them. Four reranking strategies are compared:
- chain-of-thought knowledge augmentation of a cross-attention model (CoT-KA),
- LoRA adapter tuning,
- P-tuning,
- retrieval-based ICL.

All are tested with Qwen and ChatGLM3 on the Hansel and CLEEK benchmarks. Supervised variants built on 6–7B-parameter LLMs beat the small-model baselines on every test set. Without any training, Qwen-72B nearly matches a fine-tuned Qwen-7B. The gains are largest for long-tail entities and grow with model size. An error analysis shows that most remaining mistakes concern entity granularity and entity type. The work is a practical reference for anyone linking text to knowledge bases with LLMs.

## Key points
- Supervised Qwen-7B and ChatGLM3-6B methods raise accuracy over PLM-based baselines by:
  - 3.9–11.8% on Hansel-FS,
  - 0.7–4.1% on Hansel-ZS,
  - 0.6–3.7% on CLEEK.
- Unsupervised ICL plus chain-of-thought on Qwen-72B lands within −2.4% to +1.4% of LoRA-tuned Qwen-7B.
- On long-tail entities, Qwen leads BERT by 11.8%, and the margin widens to 13.2% with more parameters. Parameter-efficient tuning overfits less than either BERT fine-tuning or full fine-tuning.
- CoT-KA encodes LLM-generated reasoning chains with a separate encoder and fuses them with cross-attention features via an MLP. It is trained with a balanced scheme of hard and random negative samples.
- Granularity errors make up 36% of failures and type errors 25%. In 56% of the granularity errors, a specific entity was linked to a more generic one.

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

Fields: [Linguistics & languages](https://md.research.elamur.ai/free/shelves/linguistics/) · [Natural language processing](https://md.research.elamur.ai/free/shelves/nlp/)

## Cite

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

```bibtex
@inproceedings{2025,
  title = {{基于大语言模型的中文实体链接实证研究}},
  author = {徐正斐 and 辛欣},
  booktitle = {自动化学报 (Acta Automatica Sinica)},
  year = {2025},
  doi = {10.16383/j.aas.c240069},
  url = {https://doi.org/10.16383/j.aas.c240069}
}
```

- [← previous 基于大语言模型智能体的代码生成综述](https://md.research.elamur.ai/free/papers/dong-2025-code-generation-llm/)
- [next → 基于大语言模型的自动代码修复综述](https://md.research.elamur.ai/free/papers/xu-2025-large-language-model/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/papers/2025/page/3/)

## References

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

1. Guo Hao, Li Xin-Yi, Tang Jiu-Yang, Guo Yan-Ming, Zhao Xiang. Adaptive feature fusion for multi-modal entity alignment. Acta Automatica Sinica, 2024, 50(4): 758−770 (郭浩, 李欣奕, 唐九阳, 郭延明, 赵翔. 自适应特征融合的多模态实 体对齐研究. 自动化学报, 2024, 50(4): 758−770) [find](https://scholar.google.com/scholar?q=Guo+Hao%2C+Li+Xin%2DYi%2C+Tang+Jiu%2DYang%2C+Guo+Yan%2DMing%2C+Zhao+Xiang%2E+Adaptive+feature+fusion+for+multi%2Dmodal+entity+alignment%2E+Acta+Automatica+Sinica%2C+2024%2C+50%284%29%3A+758%E2%88%92770+%28%E9%83%AD%E6%B5%A9%2C+%E6%9D%8E%E6%AC%A3%E5%A5%95%2C+%E5%94%90%E4%B9%9D%E9%98%B3%2C+%E9%83%AD%E5%BB%B6%E6%98%8E%2C+%E8%B5%B5%E7%BF%94%2E+%E8%87%AA%E9%80%82%E5%BA%94%E7%89%B9%E5%BE%81%E8%9E%8D%E5%90%88%E7%9A%84%E5%A4%9A%E6%A8%A1%E6%80%81%E5%AE%9E)
2. Hasibi F, Balog K, Bratsberg S E. Exploiting entity linking in queries for entity retrieval. In: Proceedings of the ACM International Conference on the Theory of Information Retrieval. Newark, Delaware, USA: Association for Computing Machinery, 2016. 209−218 [find](https://scholar.google.com/scholar?q=Hasibi+F%2C+Balog+K%2C+Bratsberg+S+E%2E+Exploiting+entity+linking+in+queries+for+entity+retrieval%2E+In%3A+Proceedings+of+the+ACM+International+Conference+on+the+Theory+of+Information+Retrieval%2E+Newark%2C+Delawar)
3. Liu Qiong-Xin, Wang Ya-Nan, Long Hang, Wang Jia-Sheng, Lu Shi-Shuai. Generative knowledge question answering technology based on global coverage mechanism and representation learning. Acta Automatica Sinica, 2022, 48(10): 2392−2405 (刘琼昕, 王亚男, 龙航, 王佳升, 卢士帅. 基于全局覆盖机制与表示 学习的生成式知识问答技术. 自动化学报, 2022, 48(1… [find](https://scholar.google.com/scholar?q=Liu+Qiong%2DXin%2C+Wang+Ya%2DNan%2C+Long+Hang%2C+Wang+Jia%2DSheng%2C+Lu+Shi%2DShuai%2E+Generative+knowledge+question+answering+technology+based+on+global+coverage+mechanism+and+representation+learning%2E+Acta+Automatica+)
4. Bai J Z, Bai S, Chu Y F, Cui Z Y, Dang K, Deng X D, et al. Qwen technical report. arXiv preprint arXiv: 2309.16609, 2023. Zeng A H, Xu B, Wang B W, Zhang C H, Yin D, Zhang D, et al. ChatGLM: A family of large language models from GLM-130B to GLM-4 all tools. arXiv preprint arXiv: 2406.12793, 2024. [find](https://scholar.google.com/scholar?q=Bai+J+Z%2C+Bai+S%2C+Chu+Y+F%2C+Cui+Z+Y%2C+Dang+K%2C+Deng+X+D%2C+et+al%2E+Qwen+technical+report%2E+arXiv+preprint+arXiv%3A+2309%2E16609%2C+2023%2E+Zeng+A+H%2C+Xu+B%2C+Wang+B+W%2C+Zhang+C+H%2C+Yin+D%2C+Zhang+D%2C+et+al%2E+ChatGLM%3A+A+family+) [our card](https://md.research.elamur.ai/free/papers/bai-2023-qwen-technical-report/)
5. Brown T B, Mann B, Ryder N, Subbiah M, Kaplan J, Dhariwal P, et al. Language models are few-shot learners. arXiv preprint arXiv: 2005.14165, 2020. [find](https://scholar.google.com/scholar?q=Brown+T+B%2C+Mann+B%2C+Ryder+N%2C+Subbiah+M%2C+Kaplan+J%2C+Dhariwal+P%2C+et+al%2E+Language+models+are+few%2Dshot+learners%2E+arXiv+preprint+arXiv%3A+2005%2E14165%2C+2020%2E)
6. Rawte V, Chakraborty S, Pathak A, Sarkar A, Tonmoy S M T I, Chadha A, et al. The troubling emergence of hallucination in large language models —— An extensive definition, quantification, and prescriptive remediations. In: Proceedings of the Conference on Empirical Methods in Natural Language Process… [find](https://scholar.google.com/scholar?q=Rawte+V%2C+Chakraborty+S%2C+Pathak+A%2C+Sarkar+A%2C+Tonmoy+S+M+T+I%2C+Chadha+A%2C+et+al%2E+The+troubling+emergence+of+hallucination+in+large+language+models+%E2%80%94%E2%80%94+An+extensive+definition%2C+quantification%2C+and+prescript)
7. Roberts A, Raffel C, Shazeer N. How much knowledge can you pack into the parameters of a language model? In: Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). Virtual Event: Association for Computational Linguistics, 2020. 5418−5426 [find](https://scholar.google.com/scholar?q=Roberts+A%2C+Raffel+C%2C+Shazeer+N%2E+How+much+knowledge+can+you+pack+into+the+parameters+of+a+language+model%3F+In%3A+Proceedings+of+the+Conference+on+Empirical+Methods+in+Natural+Language+Processing+%28EMNLP%29%2E+)
8. Kandpal N, Deng H K, Roberts A, Wallace E, Raffel C. Large language models struggle to learn long-tail knowledge. arXiv preprint arXiv: 2211.08411, 2023. [find](https://scholar.google.com/scholar?q=Kandpal+N%2C+Deng+H+K%2C+Roberts+A%2C+Wallace+E%2C+Raffel+C%2E+Large+language+models+struggle+to+learn+long%2Dtail+knowledge%2E+arXiv+preprint+arXiv%3A+2211%2E08411%2C+2023%2E) [our card](https://md.research.elamur.ai/free/papers/kandpal-2022-large-language-models-struggle-learn/)
9. Wang X T, Yang Q W, Qiu Y T, Liang J Q, He Q Y, Gu Z H, et al. KnowledGPT: Enhancing large language models with retrieval and storage access on knowledge bases. arXiv preprint arXiv: 2308.11761, 2023. — [arXiv:2308.11761](https://arxiv.org/abs/2308.11761) [find](https://scholar.google.com/scholar?q=Wang+X+T%2C+Yang+Q+W%2C+Qiu+Y+T%2C+Liang+J+Q%2C+He+Q+Y%2C+Gu+Z+H%2C+et+al%2E+KnowledGPT%3A+Enhancing+large+language+models+with+retrieval+and+storage+access+on+knowledge+bases%2E+arXiv+preprint+arXiv%3A+2308%2E11761%2C+2023%2E)
10. Ganea O-E, Hofmann T. Deep joint entity disambiguation with local neural attention. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing. Copenhagen, Denmark: Association for Computational Linguistics, 2017. 2619−2629 [find](https://scholar.google.com/scholar?q=Ganea+O%2DE%2C+Hofmann+T%2E+Deep+joint+entity+disambiguation+with+local+neural+attention%2E+In%3A+Proceedings+of+the+Conference+on+Empirical+Methods+in+Natural+Language+Processing%2E+Copenhagen%2C+Denmark%3A+Associat)
11. Le P, Titov I. Improving entity linking by modeling latent relations between mentions. In: Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics. Melbourne, Australia: Association for Computational Linguistics, 2018. 1595−1604 [find](https://scholar.google.com/scholar?q=Le+P%2C+Titov+I%2E+Improving+entity+linking+by+modeling+latent+relations+between+mentions%2E+In%3A+Proceedings+of+the+56th+Annual+Meeting+of+the+Association+for+Computational+Linguistics%2E+Melbourne%2C+Australia)
12. Jiang Sheng-Chen, Wang Hong-Bin, Yu Zheng-Tao, Xian Yan-Tuan, Wang Hong-Tao. Domain integrated-entity links based on relationship indices and representation learning. Acta Automatica Sinica, 2021, 47(10): 2376−2385 (蒋胜臣, 王红斌, 余正涛, 线岩团, 王红涛. 基于关系指数和表示学 习的领域集成实体链接. 自动化学报, 2021, 47(10): 2376−2385) [find](https://scholar.google.com/scholar?q=Jiang+Sheng%2DChen%2C+Wang+Hong%2DBin%2C+Yu+Zheng%2DTao%2C+Xian+Yan%2DTuan%2C+Wang+Hong%2DTao%2E+Domain+integrated%2Dentity+links+based+on+relationship+indices+and+representation+learning%2E+Acta+Automatica+Sinica%2C+2021%2C+47%28)
13. Devlin J, Chang M W, Lee K, Toutanova K. BERT: Pre-training of deep bidirectional transformers for language understanding. In: Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers).… [find](https://scholar.google.com/scholar?q=Devlin+J%2C+Chang+M+W%2C+Lee+K%2C+Toutanova+K%2E+BERT%3A+Pre%2Dtraining+of+deep+bidirectional+transformers+for+language+understanding%2E+In%3A+Proceedings+of+the+Conference+of+the+North+American+Chapter+of+the+Associ)
14. Yamada I, Washio K, Shindo H, Matsumoto Y. Global entity disambiguation with BERT. In: Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Seattle, USA: Association for Computational Linguistics, 2022. 3264−3271 [find](https://scholar.google.com/scholar?q=Yamada+I%2C+Washio+K%2C+Shindo+H%2C+Matsumoto+Y%2E+Global+entity+disambiguation+with+BERT%2E+In%3A+Proceedings+of+the+Conference+of+the+North+American+Chapter+of+the+Association+for+Computational+Linguistics%3A+Hum)
15. Wu L, Petroni F, Josifoski M, Riedel S, Zettlemoyer L. Scalable zero-shot entity linking with dense entity retrieval. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP). Virtual Event: Association for Computational Linguistics, 2020. 6397−6407 [find](https://scholar.google.com/scholar?q=Wu+L%2C+Petroni+F%2C+Josifoski+M%2C+Riedel+S%2C+Zettlemoyer+L%2E+Scalable+zero%2Dshot+entity+linking+with+dense+entity+retrieval%2E+In%3A+Proceedings+of+the+Conference+on+Empirical+Methods+in+Natural+Language+Process) [our card](https://md.research.elamur.ai/free/papers/wu-2020-scalable-zero-shot-entity-linking/)
16. Wei J, Wang X Z, Schuurmans D, Bosma M, Ichter B, Xia F, et al. Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 2022, 35: 24824−24837 [find](https://scholar.google.com/scholar?q=Wei+J%2C+Wang+X+Z%2C+Schuurmans+D%2C+Bosma+M%2C+Ichter+B%2C+Xia+F%2C+et+al%2E+Chain%2Dof%2Dthought+prompting+elicits+reasoning+in+large+language+models%2E+Advances+in+Neural+Information+Processing+Systems%2C+2022%2C+35%3A+2482) [our card](https://md.research.elamur.ai/free/papers/wei-2022-chain-thought-prompting-elicits-reasoning/)
17. Ouyang L, Wu J, Jiang X, Almeida D, Wainwright C, Mishkin P, et al. Training language models to follow instructions with human feedback. arXiv preprint arXiv: 2203.02155, 2022. [find](https://scholar.google.com/scholar?q=Ouyang+L%2C+Wu+J%2C+Jiang+X%2C+Almeida+D%2C+Wainwright+C%2C+Mishkin+P%2C+et+al%2E+Training+language+models+to+follow+instructions+with+human+feedback%2E+arXiv+preprint+arXiv%3A+2203%2E02155%2C+2022%2E)
18. Talmor A, Tafjord O, Clark P, Goldberg Y, Berant J. Leap-of-thought: Teaching pre-trained models to systematically reason over implicit knowledge. Advances in Neural Information Processing Systems, 2020, 33: 20227−20237 [find](https://scholar.google.com/scholar?q=Talmor+A%2C+Tafjord+O%2C+Clark+P%2C+Goldberg+Y%2C+Berant+J%2E+Leap%2Dof%2Dthought%3A+Teaching+pre%2Dtrained+models+to+systematically+reason+over+implicit+knowledge%2E+Advances+in+Neural+Information+Processing+Systems%2C+20)
19. Xu Z R, Shan Z F, Li Y X, Hu B T, Qin B. Hansel: A Chinese few-shot and zero-shot entity linking benchmark. In: Proceedings of the 16th ACM International Conference on Web Search and Data Mining. Singapore: Association for Computing Machinery, 2023. 832−840 [find](https://scholar.google.com/scholar?q=Xu+Z+R%2C+Shan+Z+F%2C+Li+Y+X%2C+Hu+B+T%2C+Qin+B%2E+Hansel%3A+A+Chinese+few%2Dshot+and+zero%2Dshot+entity+linking+benchmark%2E+In%3A+Proceedings+of+the+16th+ACM+International+Conference+on+Web+Search+and+Data+Mining%2E+Sing)
20. Zeng W X, Zhao X, Tang J Y, Tan Z, Huang X Q. CLEEK: A Chinese long-text corpus for entity linking. In: Proceedings of the 12th Language Resources and Evaluation Conference. Marseille, France: European Language Resources Association, 2020. 2026−2035 [find](https://scholar.google.com/scholar?q=Zeng+W+X%2C+Zhao+X%2C+Tang+J+Y%2C+Tan+Z%2C+Huang+X+Q%2E+CLEEK%3A+A+Chinese+long%2Dtext+corpus+for+entity+linking%2E+In%3A+Proceedings+of+the+12th+Language+Resources+and+Evaluation+Conference%2E+Marseille%2C+France%3A+Europea)
21. Wu D J, Zhang J, Huang X M. Chain of thought prompting elicits knowledge augmentation. arXiv preprint arXiv: 2307.1640, 2023. [find](https://scholar.google.com/scholar?q=Wu+D+J%2C+Zhang+J%2C+Huang+X+M%2E+Chain+of+thought+prompting+elicits+knowledge+augmentation%2E+arXiv+preprint+arXiv%3A+2307%2E1640%2C+2023%2E)
22. Humeau S, Shuster K, Lachaux M A, Weston J. Poly-encoders: Transformer architectures and pre-training strategies for fast and accurate multi-sentence scoring. arXiv preprint arXiv: 1905.01969, 2020. — [arXiv:1905.01969](https://arxiv.org/abs/1905.01969) [find](https://scholar.google.com/scholar?q=Humeau+S%2C+Shuster+K%2C+Lachaux+M+A%2C+Weston+J%2E+Poly%2Dencoders%3A+Transformer+architectures+and+pre%2Dtraining+strategies+for+fast+and+accurate+multi%2Dsentence+scoring%2E+arXiv+preprint+arXiv%3A+1905%2E01969%2C+2020%2E+%E2%80%94)
23. Houlsby N, Giurgiu A, Jastrzebski S, Morrone B, Laroussilhe Q D, Gesmundo A, et al. Parameter-efficient transfer learning for NLP. arXiv preprint arXiv: 1902.00751, 2019. [find](https://scholar.google.com/scholar?q=Houlsby+N%2C+Giurgiu+A%2C+Jastrzebski+S%2C+Morrone+B%2C+Laroussilhe+Q+D%2C+Gesmundo+A%2C+et+al%2E+Parameter%2Defficient+transfer+learning+for+NLP%2E+arXiv+preprint+arXiv%3A+1902%2E00751%2C+2019%2E)
24. Hu E J, Shen Y L, Wallis P, Allen-Zhu Z Y, Li Y Z, Wang S, et al. LoRA: Low-rank adaptation of large language models. arXiv preprint arXiv: 2106.09685, 2022. [find](https://scholar.google.com/scholar?q=Hu+E+J%2C+Shen+Y+L%2C+Wallis+P%2C+Allen%2DZhu+Z+Y%2C+Li+Y+Z%2C+Wang+S%2C+et+al%2E+LoRA%3A+Low%2Drank+adaptation+of+large+language+models%2E+arXiv+preprint+arXiv%3A+2106%2E09685%2C+2022%2E)
25. Liu X, Zheng Y N, Du Z X, Ding M, Qian Y J, Yang Z L, et al. GPT understands, too. AI Open, 2024, 5: 208−215 [find](https://scholar.google.com/scholar?q=Liu+X%2C+Zheng+Y+N%2C+Du+Z+X%2C+Ding+M%2C+Qian+Y+J%2C+Yang+Z+L%2C+et+al%2E+GPT+understands%2C+too%2E+AI+Open%2C+2024%2C+5%3A+208%E2%88%92215)
26. Liu J C, Shen D H, Zhang Y Z, Dolan B, Carin L, Chen W Z. What makes good in-context examples for GPT-3? In: Proceedings of the Deep Learning Inside Out (DeeLIO 2022): The 3rd Workshop on Knowledge Extraction and Integration for Deep Learning Architectures. Dublin, Ireland: Association for Computati… [find](https://scholar.google.com/scholar?q=Liu+J+C%2C+Shen+D+H%2C+Zhang+Y+Z%2C+Dolan+B%2C+Carin+L%2C+Chen+W+Z%2E+What+makes+good+in%2Dcontext+examples+for+GPT%2D3%3F+In%3A+Proceedings+of+the+Deep+Learning+Inside+Out+%28DeeLIO+2022%29%3A+The+3rd+Workshop+on+Knowledge+Ex) [our card](https://md.research.elamur.ai/free/papers/liu-2022-what-makes-good-context-examples/)
27. Reimers N, Gurevych I. Sentence-BERT: Sentence embeddings using siamese BERT-networks. arXiv preprint arXiv: 1908.10084, 2019. [find](https://scholar.google.com/scholar?q=Reimers+N%2C+Gurevych+I%2E+Sentence%2DBERT%3A+Sentence+embeddings+using+siamese+BERT%2Dnetworks%2E+arXiv+preprint+arXiv%3A+1908%2E10084%2C+2019%2E)
28. Huang S J, Wang B B, Qin L B, Zhao Q, Xu R F. Improving few-shot and zero-shot entity linking with coarse-to-fine lexicon-based retriever. In: Processings of the Natural Language Processing and Chinese Computing: 12th National CCF Conference. Foshan, China: 2023. 245−256 [find](https://scholar.google.com/scholar?q=Huang+S+J%2C+Wang+B+B%2C+Qin+L+B%2C+Zhao+Q%2C+Xu+R+F%2E+Improving+few%2Dshot+and+zero%2Dshot+entity+linking+with+coarse%2Dto%2Dfine+lexicon%2Dbased+retriever%2E+In%3A+Processings+of+the+Natural+Language+Processing+and+Chines)
29. Zhou H Y, Sun C J, Lin L, Shan L L. ERNIE-AT-CEL: A Chinese few-shot emerging entity linking model based on ERNIE and adversarial training. In: Processings of the Natural Language Processing and Chinese Computing: 12th National CCF Conference. Foshan, China: 2023. 48−56 [find](https://scholar.google.com/scholar?q=Zhou+H+Y%2C+Sun+C+J%2C+Lin+L%2C+Shan+L+L%2E+ERNIE%2DAT%2DCEL%3A+A+Chinese+few%2Dshot+emerging+entity+linking+model+based+on+ERNIE+and+adversarial+training%2E+In%3A+Processings+of+the+Natural+Language+Processing+and+Chine)
30. Xu Z, Shan Z, Hu B, Zhang M. Overview of the NLPCC 2023 shared task 6: Chinese few-shot and zero-shot entity linking. In: Proceedings of Natural Language Processing and Chinese Computing: 12th National CCF Conference. [find](https://scholar.google.com/scholar?q=Xu+Z%2C+Shan+Z%2C+Hu+B%2C+Zhang+M%2E+Overview+of+the+NLPCC+2023+shared+task+6%3A+Chinese+few%2Dshot+and+zero%2Dshot+entity+linking%2E+In%3A+Proceedings+of+Natural+Language+Processing+and+Chinese+Computing%3A+12th+Nationa)
31. Foshan, China: 2023. 257−265 de Cao N, Wu L, Popat K, Artetxe M, Goyal N, Plekhanov M, et al. Multilingual autoregressive entity linking. Transactions of the Association for Computational Linguistics, 2022, 10: 274−290 [find](https://scholar.google.com/scholar?q=Foshan%2C+China%3A+2023%2E+257%E2%88%92265+de+Cao+N%2C+Wu+L%2C+Popat+K%2C+Artetxe+M%2C+Goyal+N%2C+Plekhanov+M%2C+et+al%2E+Multilingual+autoregressive+entity+linking%2E+Transactions+of+the+Association+for+Computational+Linguistics%2C)
32. Zhang Q R, Chen M S, Bukharin A, He P C, Cheng Y, Chen W Z, et al. Adaptive budget allocation for parameter-efficient fine-tuning. arXiv preprint arXiv: 2303.10512v1, 2023. [find](https://scholar.google.com/scholar?q=Zhang+Q+R%2C+Chen+M+S%2C+Bukharin+A%2C+He+P+C%2C+Cheng+Y%2C+Chen+W+Z%2C+et+al%2E+Adaptive+budget+allocation+for+parameter%2Defficient+fine%2Dtuning%2E+arXiv+preprint+arXiv%3A+2303%2E10512v1%2C+2023%2E) [our card](https://md.research.elamur.ai/free/papers/zhang-2023-adalora-adaptive-budget-allocation-parameter/)
33. Wei J, Tay Y, Bommasani R, Raffel C, Zoph B, Borgeaud S, et al. Emergent abilities of large language models [Online], available: [https://openreview.net/forum?id=yzkSU5zdwD](https://openreview.net/forum?id=yzkSU5zdwD), January 15, 2025 [↗](https://openreview.net/forum?id=yzkSU5zdwD)
34. Bonifacio L, Abonizio H, Fadaee M, Nogueira R. InPars: Unsupervised dataset generation for information retrieval. In: Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. Madrid, Spain: Association for Computing Machinery, 2022. 2387−2392 [find](https://scholar.google.com/scholar?q=Bonifacio+L%2C+Abonizio+H%2C+Fadaee+M%2C+Nogueira+R%2E+InPars%3A+Unsupervised+dataset+generation+for+information+retrieval%2E+In%3A+Proceedings+of+the+45th+International+ACM+SIGIR+Conference+on+Research+and+Develop)
35. Ferraretto F, Laitz T, Lotufo R, Nogueira R. ExaRanker: Synthetic explanations improve neural rankers. In: Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. Taipei, China: Association for Computing Machinery, 2023. 2409−2414 [find](https://scholar.google.com/scholar?q=Ferraretto+F%2C+Laitz+T%2C+Lotufo+R%2C+Nogueira+R%2E+ExaRanker%3A+Synthetic+explanations+improve+neural+rankers%2E+In%3A+Proceedings+of+the+46th+International+ACM+SIGIR+Conference+on+Research+and+Development+in+Inf)
36. Dettmers T, Pagnoni A, Holtzman A, Zettlemoyer L. QLoRA: Efficient finetuning of quantized LLMs. arXiv preprint arXiv: 2305.14314, 2023. [find](https://scholar.google.com/scholar?q=Dettmers+T%2C+Pagnoni+A%2C+Holtzman+A%2C+Zettlemoyer+L%2E+QLoRA%3A+Efficient+finetuning+of+quantized+LLMs%2E+arXiv+preprint+arXiv%3A+2305%2E14314%2C+2023%2E)
37. Li X L, Liang P. Prefix-tuning: Optimizing continuous prompts for generation. In: Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing. Virtual Event: Association for Computational Linguisti… [find](https://scholar.google.com/scholar?q=Li+X+L%2C+Liang+P%2E+Prefix%2Dtuning%3A+Optimizing+continuous+prompts+for+generation%2E+In%3A+Proceedings+of+the+59th+Annual+Meeting+of+the+Association+for+Computational+Linguistics+and+the+11th+International+Joi)
38. Lester B, Al-Rfou R, Constant N. The power of scale for parameter-efficient prompt tuning. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing. Virtual Event: Punta Cana, Dominican Republic: Association for Computational Linguistics, 2021. 3045−3059 [find](https://scholar.google.com/scholar?q=Lester+B%2C+Al%2DRfou+R%2C+Constant+N%2E+The+power+of+scale+for+parameter%2Defficient+prompt+tuning%2E+In%3A+Proceedings+of+the+Conference+on+Empirical+Methods+in+Natural+Language+Processing%2E+Virtual+Event%3A+Punta+C)
39. Liu X, Ji K X, Fu Y C, Tam W L, Du Z X, Yang Z L, et al. P-Tuning: Prompt tuning can be comparable to fine-tuning across scales and tasks. In: Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers). Dublin, Ireland: Association for Computatio… [find](https://scholar.google.com/scholar?q=Liu+X%2C+Ji+K+X%2C+Fu+Y+C%2C+Tam+W+L%2C+Du+Z+X%2C+Yang+Z+L%2C+et+al%2E+P%2DTuning%3A+Prompt+tuning+can+be+comparable+to+fine%2Dtuning+across+scales+and+tasks%2E+In%3A+Proceedings+of+the+60th+Annual+Meeting+of+the+Association) [our card](https://md.research.elamur.ai/free/papers/liu-2022-p-tuning-prompt-tuning-can/)
40. Lu Y, Bartolo M, Moore A, Riedel S, Stenetorp P. Fantastically ordered prompts and where to find them: Overcoming few-shot prompt order sensitivity. In: Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics. Dublin, Ireland: Association for Computational Linguistics… [find](https://scholar.google.com/scholar?q=Lu+Y%2C+Bartolo+M%2C+Moore+A%2C+Riedel+S%2C+Stenetorp+P%2E+Fantastically+ordered+prompts+and+where+to+find+them%3A+Overcoming+few%2Dshot+prompt+order+sensitivity%2E+In%3A+Proceedings+of+the+60th+Annual+Meeting+of+the+A) [our card](https://md.research.elamur.ai/free/papers/lu-2022-fantastically-ordered-prompts-where-find/)
41. Zhao T, Wallace E, Feng S, Klein D, Singh S. Calibrate before use: Improving few-shot performance of language models. arXiv preprint arXiv: 2102.09690, 2021. [find](https://scholar.google.com/scholar?q=Zhao+T%2C+Wallace+E%2C+Feng+S%2C+Klein+D%2C+Singh+S%2E+Calibrate+before+use%3A+Improving+few%2Dshot+performance+of+language+models%2E+arXiv+preprint+arXiv%3A+2102%2E09690%2C+2021%2E) [our card](https://md.research.elamur.ai/free/papers/zhao-2021-calibrate-before-use-improving-few/)
42. Rubin O, Herzig J, Berant J. Learning to retrieve prompts for in-context learning. In: Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Seattle, USA: Association for Computational Linguistics, 2022. 2655−2671 [find](https://scholar.google.com/scholar?q=Rubin+O%2C+Herzig+J%2C+Berant+J%2E+Learning+to+retrieve+prompts+for+in%2Dcontext+learning%2E+In%3A+Proceedings+of+the+Conference+of+the+North+American+Chapter+of+the+Association+for+Computational+Linguistics%3A+Hum) [our card](https://md.research.elamur.ai/free/papers/rubin-2021-learning-retrieve-prompts-context-learning/)
43. Min S, Lewis M, Zettlemoyer L, Hajishirzi H. MetaICL: Learning to learn in context. In: Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Seattle, USA: Association for Computational Linguistics, 2022. 2791−2809 [find](https://scholar.google.com/scholar?q=Min+S%2C+Lewis+M%2C+Zettlemoyer+L%2C+Hajishirzi+H%2E+MetaICL%3A+Learning+to+learn+in+context%2E+In%3A+Proceedings+of+the+Conference+of+the+North+American+Chapter+of+the+Association+for+Computational+Linguistics%3A+Hu)
44. Cho Y M, Zhang L, Callison-Burch C. Unsupervised entity linking with guided summarization and multiple-choice selection. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing. Abu Dhabi, United Arab Emirates: Association for Computational Linguistics, 2022. 9394−9401 [find](https://scholar.google.com/scholar?q=Cho+Y+M%2C+Zhang+L%2C+Callison%2DBurch+C%2E+Unsupervised+entity+linking+with+guided+summarization+and+multiple%2Dchoice+selection%2E+In%3A+Proceedings+of+the+Conference+on+Empirical+Methods+in+Natural+Language+Proc)
45. Shi S B, Xu Z R, Hu B T, Zhang M. Generative multimodal entity linking. In: Proceedings of the Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). [find](https://scholar.google.com/scholar?q=Shi+S+B%2C+Xu+Z+R%2C+Hu+B+T%2C+Zhang+M%2E+Generative+multimodal+entity+linking%2E+In%3A+Proceedings+of+the+Joint+International+Conference+on+Computational+Linguistics%2C+Language+Resources+and+Evaluation+%28LREC%2DCOLI)
46. Torino, Italy: ELRA and ICCL, 2024. 7654−7665 de Cao N, Izacard G, Riedel S, Petroni F. Autoregressive entity retrieval. In: Proceedings of the 9th International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Torino%2C+Italy%3A+ELRA+and+ICCL%2C+2024%2E+7654%E2%88%927665+de+Cao+N%2C+Izacard+G%2C+Riedel+S%2C+Petroni+F%2E+Autoregressive+entity+retrieval%2E+In%3A+Proceedings+of+the+9th+International+Conference+on+Learning+Representations)
47. Virtual Event: ICLR, 2021. [find](https://scholar.google.com/scholar?q=Virtual+Event%3A+ICLR%2C+2021%2E)
48. 徐正斐 北京理工大学计算机学院硕 士研究生. 主要研究方向为知识工程. E-mail: zhengfei@bit.edu.cn (XU Zheng-Fei Master student at the School of Computer Science and Technology, Beijing Institute of Technology. His main research interest is knowledge engineering.) 辛 欣 北京理工大学计算机学院副 教授. 主要研究方向为自然语言处理, 知识工程, 信息检索. 本文通信作者. E-mail: xxin@b… [find](https://scholar.google.com/scholar?q=%E5%BE%90%E6%AD%A3%E6%96%90+%E5%8C%97%E4%BA%AC%E7%90%86%E5%B7%A5%E5%A4%A7%E5%AD%A6%E8%AE%A1%E7%AE%97%E6%9C%BA%E5%AD%A6%E9%99%A2%E7%A1%95+%E5%A3%AB%E7%A0%94%E7%A9%B6%E7%94%9F%2E+%E4%B8%BB%E8%A6%81%E7%A0%94%E7%A9%B6%E6%96%B9%E5%90%91%E4%B8%BA%E7%9F%A5%E8%AF%86%E5%B7%A5%E7%A8%8B%2E+E%2Dmail%3A+zhengfei%40bit%2Eedu%2Ecn+%28XU+Zheng%2DFei+Master+student+at+the+School+of+Computer+Science+and+Technology%2C+Beijing+Institute+of+Technology%2E+His+main+research+intere)