---
title: "PretrainRL — paper by Liu et al. (2026) — summary and ideas"
description: "Rather than teaching refusal (\"I don't know\") or applying post-hoc knowledge editing, the authors introduce PretrainRL. 56 references."
authors: ["Langming Liu", "Kangtao Lv", "Haibin Chen", "Weidong Zhang", "Yejing Wang", "Shilei Liu", "Xin Tong", "Yujin Yuan", "Yongwei Wang", "Wenbo Su", "Bo Zheng"]
year: 2026
venue: "Findings of the Association for Computational Linguistics: ACL 2026"
source: "https://aclanthology.org/2026.findings-acl.910/"
doi: "10.18653/v1/2026.findings-acl.910"
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/liu-2026-pretrainrl-alleviating-factuality-hallucination-large/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/liu-2026-pretrainrl-alleviating-factuality-hallucination-large/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/39/) / PretrainRL: Alleviating Factuality Hallucination of Large Language Models at the Beginning

# PretrainRL: Alleviating Factuality Hallucination of Large Language Models at the Beginning

**Authors:** [Langming Liu](https://md.research.elamur.ai/free/en/authors/langming-liu/), [Kangtao Lv](https://md.research.elamur.ai/free/en/authors/kangtao-lv/), [Haibin Chen](https://md.research.elamur.ai/free/en/authors/haibin-chen/), [Weidong Zhang](https://md.research.elamur.ai/free/en/authors/weidong-zhang/), Yejing Wang, [Shilei Liu](https://md.research.elamur.ai/free/en/authors/shilei-liu/), [Xin Tong](https://md.research.elamur.ai/free/en/authors/xin-tong/), [Yujin Yuan](https://md.research.elamur.ai/free/en/authors/yujin-yuan/), Yongwei Wang, [Wenbo Su](https://md.research.elamur.ai/free/en/authors/wenbo-su/) and [Bo Zheng](https://md.research.elamur.ai/free/en/authors/bo-zheng/) · **Year:** 2026 · **Venue:** [Findings of the Association for Computational Linguistics: ACL 2026](https://md.research.elamur.ai/free/en/venues/findings/) ([Findings of the Association for Computational Linguistics 2026](https://md.research.elamur.ai/free/en/venues/findings/2026/)) · **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/2026.findings-acl.910/) · [PDF](https://aclanthology.org/2026.findings-acl.910.pdf) · [DOI 10.18653/v1/2026.findings-acl.910](https://doi.org/10.18653/v1/2026.findings-acl.910) · [OpenAlex](https://openalex.org/W7166869576)

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

## Summary
The paper addresses factual hallucinations in LLMs, meaning confident statements that can be checked and shown false. The authors trace them to frequency imbalance in pretraining data. Under next-token prediction, popular ("head") facts can dominate rare ("tail") ones. The model may then disregard the subject and predict from the relation alone, so correct tail answers end up with very low probability.

Rather than teaching refusal ("I don't know") or applying post-hoc knowledge editing, the authors introduce PretrainRL. It adds reinforcement learning to the pretraining stage, following a "debiasing then learning" idea. Using DPO, the model first lowers the probability of its favoured wrong answers and then raises the probability of the true object in a knowledge triple. Wrong answers to reject are mined with beam search, and new metrics describe the model's probability state over facts.

On three public benchmarks, including POPQA, PretrainRL reduces factual hallucinations and is reported to beat strong LLMs. General abilities are largely preserved, whereas SFT damages them badly.

## Key points
- **Negative sampling:** the model is prompted with the subject and predicate, and beam search produces candidate answers. Candidates are ranked by the product of their token probabilities, and the gold answer is removed. The top candidates (e.g., 5 or 10) become DPO rejected samples, and they are pooled per question category for efficiency.
- **Models and comparison:** the method is applied to Qwen3-4B and Qwen3-8B. On POPQA (pass@1) they are compared with GPT-4o, Claude-sonnet and DeepSeek-v3/R1. The extracted figure is too garbled to attribute exact scores reliably.
- **General capability check (Qwen3-4B, OpenCompass), scores given as base / PretrainRL / SFT:**

  | Benchmark | Base | PretrainRL | SFT |
  |---|---|---|---|
  | CEval | 50.81 | 49.06 | 29.38 |
  | MMLU | 52.39 | 52.01 | 31.76 |
  | MATH | 14.46 | 17.22 | 0.26 |
  | GSM8K | 83.62 | 80.97 | 0.53 |
  | BBH | 71.80 | 66.07 | 4.02 |

  PretrainRL changes little on most benchmarks, though BBH drops by about 6 points; SFT collapses on all of them.
- **Case study:** for "What is Yungay the capital of?", the correct answer "Yungay Province" ranked only 29th among the top 50 beam candidates, with probability ≈5.6e-33. This suggests the fact is stored in the model but heavily outweighed by wrong answers.
- **Limitation:** the sampling relies on the dataset having distinct question categories. The authors suggest alternatives such as RLVR when it does not.

*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

- [Alibaba Group](https://md.research.elamur.ai/free/en/companies/alibaba-group/) — Company, Hangzhou, China · 168 works · 15,386 citations
- [Zhejiang University](https://md.research.elamur.ai/free/en/universities/zhejiang-university/) — University, Hangzhou, China · 172 works · 12,988 citations
- [Shanghai Artificial Intelligence Laboratory](https://md.research.elamur.ai/free/en/labs/shanghai-artificial-intelligence-laboratory/) — Research institute, Shanghai, China · 117 works · 11,281 citations
- [Alibaba Group (United States)](https://md.research.elamur.ai/free/en/companies/alibaba-group-united-states/) — Company, San Mateo, United States · 7 works · 329 citations

📚 Shelves (5): [🖥️ 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 · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 2,869 works

🧭 Topics (1): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works

## Cite

Cite as: Langming Liu, Kangtao Lv, Haibin Chen, Weidong Zhang, Yejing Wang, Shilei Liu, Xin Tong, Yujin Yuan, Yongwei Wang, Wenbo Su and Bo Zheng. 2026. PretrainRL: Alleviating Factuality Hallucination of Large Language Models at the Beginning. Findings of the Association for Computational Linguistics: ACL 2026.

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

```bibtex
@inproceedings{liu2026pretrainrl,
  title = {{PretrainRL: Alleviating Factuality Hallucination of Large Language Models at the Beginning}},
  author = {Langming Liu, Kangtao Lv, Haibin Chen, Weidong Zhang, Yejing Wang, Shilei Liu, Xin Tong, Yujin Yuan, Yongwei Wang, Wenbo Su and Bo Zheng},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.910},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.910}
}
```

- [← previous PressMint QuickCheck: Operationalising Readiness Diagnostics for Interoperable Historical Newspaper Corpora](https://md.research.elamur.ai/free/en/papers/moro-2026-pressmint-quickcheck-operationalising-readiness-diagnostics/)
- [next → Preventing Safety Drift in Large Language Models via Coupled Weight and Activation Constraints](https://md.research.elamur.ai/free/en/papers/peng-2026-preventing-safety-drift-large-language/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/39/)

## References

56 references, 24 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Yuntao Bai, Andy Jones, Kamal Ndousse, Amanda Askell, Anna Chen, Nova DasSarma, Dawn Drain, Stanislav Fort, Deep Ganguli, Tom Henighan, and 1 others. 2022. Training a helpful and harmless assistant with reinforcement learning from human feedback. arXiv preprint arXiv:2204.05862. [find](https://scholar.google.com/scholar?q=Yuntao+Bai%2C+Andy+Jones%2C+Kamal+Ndousse%2C+Amanda+Askell%2C+Anna+Chen%2C+Nova+DasSarma%2C+Dawn+Drain%2C+Stanislav+Fort%2C+Deep+Ganguli%2C+Tom+Henighan%2C+and+1+others%2E+2022%2E+Training+a+helpful+and+harmless+assistant+wi)
2. Lang Cao. 2023. Learn to refuse: Making large language models more controllable and reliable through knowledge scope limitation and refusal mechanism. arXiv preprint arXiv:2311.01041. [find](https://scholar.google.com/scholar?q=Lang+Cao%2E+2023%2E+Learn+to+refuse%3A+Making+large+language+models+more+controllable+and+reliable+through+knowledge+scope+limitation+and+refusal+mechanism%2E+arXiv+preprint+arXiv%3A2311%2E01041%2E)
3. Nicola De Cao, Wilker Aziz, and Ivan Titov. 2021. Editing factual knowledge in language models. Preprint, arXiv:2104.08164. [find](https://scholar.google.com/scholar?q=Nicola+De+Cao%2C+Wilker+Aziz%2C+and+Ivan+Titov%2E+2021%2E+Editing+factual+knowledge+in+language+models%2E+Preprint%2C+arXiv%3A2104%2E08164%2E)
4. François Charton and Julia Kempe. 2024. Emergent properties with repeated examples. Preprint, arXiv:2410.07041. — [arXiv:2410.07041](https://arxiv.org/abs/2410.07041) [find](https://scholar.google.com/scholar?q=Fran%C3%A7ois+Charton+and+Julia+Kempe%2E+2024%2E+Emergent+properties+with+repeated+examples%2E+Preprint%2C+arXiv%3A2410%2E07041%2E+%E2%80%94+%5BarXiv%3A2410%2E07041%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2410%2E07041%29)
5. Zhe Chen, Yuehan Wang, Bin Zhao, Jing Cheng, Xin Zhao, and Zongtao Duan. 2020. Knowledge graph completion: A review. Ieee Access, 8:192435– 192456. [find](https://scholar.google.com/scholar?q=Zhe+Chen%2C+Yuehan+Wang%2C+Bin+Zhao%2C+Jing+Cheng%2C+Xin+Zhao%2C+and+Zongtao+Duan%2E+2020%2E+Knowledge+graph+completion%3A+A+review%2E+Ieee+Access%2C+8%3A192435%E2%80%93+192456%2E)
6. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. 2021. [Training verifiers to solve math word problems](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/). arXiv preprint arXiv:2110.14168.
7. OpenCompass Contributors. 2023. Opencompass: A universal evaluation platform for foundation models. [https://github.com/open-compass/](https://github.com/open-compass/) opencompass. [↗](https://github.com/open-compass/)
8. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, and 1 others. 2024. [The llama 3 herd of models](https://md.research.elamur.ai/free/en/papers/grattafiori-2024-llama-3-herd-models/). arXiv e-prints, pages arXiv–2407.
9. Leo Gao, Stella Biderman, Sid Black, Laurence Golding, Travis Hoppe, Charles Foster, Jason Phang, Horace He, Anish Thite, Noa Nabeshima, and 1 others. 2020. The pile: An 800gb dataset of diverse text for language modeling. arXiv preprint arXiv:2101.00027. [find](https://scholar.google.com/scholar?q=Leo+Gao%2C+Stella+Biderman%2C+Sid+Black%2C+Laurence+Golding%2C+Travis+Hoppe%2C+Charles+Foster%2C+Jason+Phang%2C+Horace+He%2C+Anish+Thite%2C+Noa+Nabeshima%2C+and+1+others%2E+2020%2E+The+pile%3A+An+800gb+dataset+of+diverse+text+)
10. Zorik Gekhman, Gal Yona, Roee Aharoni, Matan Eyal, Amir Feder, Roi Reichart, and Jonathan Herzig. 2024. Does fine-tuning LLMs on new knowledge encourage hallucinations? In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 7765–7784, Miami, Florida, USA. As… [find](https://scholar.google.com/scholar?q=Zorik+Gekhman%2C+Gal+Yona%2C+Roee+Aharoni%2C+Matan+Eyal%2C+Amir+Feder%2C+Roi+Reichart%2C+and+Jonathan+Herzig%2E+2024%2E+Does+fine%2Dtuning+LLMs+on+new+knowledge+encourage+hallucinations%3F+In+Proceedings+of+the+2024+Conf)
11. Aaron Grattafiori and Abhimanyu Dubey et al. 2024. [The llama 3 herd of models](https://md.research.elamur.ai/free/en/papers/grattafiori-2024-llama-3-herd-models/). Preprint, arXiv:2407.21783.
12. Arnav Gudibande, Eric Wallace, Charlie Snell, Xinyang Geng, Hao Liu, Pieter Abbeel, Sergey Levine, and Dawn Song. 2023. [The false promise of imitating proprietary llms](https://md.research.elamur.ai/free/en/papers/gudibande-2023-false-promise-imitating-proprietary-llms/). Preprint, arXiv:2305.15717.
13. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2021. [Measuring massive multitask language understanding](https://md.research.elamur.ai/free/en/papers/hendrycks-2020-measuring-massive-multitask-language-understanding/). Proceedings of the International Conference on Learning Representations (ICLR).
14. Aidan Hogan, Eva Blomqvist, Michael Cochez, Claudia d’Amato, Gerard De Melo, Claudio Gutierrez, Sabrina Kirrane, José Emilio Labra Gayo, Roberto Navigli, Sebastian Neumaier, and 1 others. 2021. [Knowledge graphs](https://md.research.elamur.ai/free/en/papers/hogan-2020-knowledge-graphs/). ACM Computing Surveys (Csur), 54(4):1–37.
15. Yuzhen Huang, Yuzhuo Bai, Zhihao Zhu, Junlei Zhang, Jinghan Zhang, Tangjun Su, Junteng Liu, Chuancheng Lv, Yikai Zhang, Jiayi Lei, Yao Fu, Maosong Sun, and Junxian He. 2023. [C-eval: A multi-level multi-discipline chinese evaluation suite for foundation models](https://md.research.elamur.ai/free/en/papers/huang-2023-c-eval-multi-level-multi/). In Advances in Neural Information Proce…
16. Ziwei Ji, Nayeon Lee, Rita Frieske, Tiezheng Yu, Dan Su, Yan Xu, Etsuko Ishii, Ye Jin Bang, Andrea Madotto, and Pascale Fung. 2023. [Survey of hallucination in natural language generation](https://md.research.elamur.ai/free/en/papers/ji-2022-survey-hallucination-natural-language-generation/). ACM computing surveys, 55(12):1–38.
17. DeepSeek-AI. 2024. Deepseek-v3 technical report. Preprint, arXiv:2412.19437. [find](https://scholar.google.com/scholar?q=DeepSeek%2DAI%2E+2024%2E+Deepseek%2Dv3+technical+report%2E+Preprint%2C+arXiv%3A2412%2E19437%2E)
18. Saurav Kadavath, Tom Conerly, Amanda Askell, Tom Henighan, Dawn Drain, Ethan Perez, Nicholas Schiefer, Zac Hatfield-Dodds, Nova DasSarma, Eli Tran-Johnson, and 1 others. 2022. Language models (mostly) know what they know. arXiv preprint arXiv:2207.05221. [find](https://scholar.google.com/scholar?q=Saurav+Kadavath%2C+Tom+Conerly%2C+Amanda+Askell%2C+Tom+Henighan%2C+Dawn+Drain%2C+Ethan+Perez%2C+Nicholas+Schiefer%2C+Zac+Hatfield%2DDodds%2C+Nova+DasSarma%2C+Eli+Tran%2DJohnson%2C+and+1+others%2E+2022%2E+Language+models+%28mostly%29)
19. Qingxiu Dong, Li Dong, Yao Tang, Tianzhu Ye, Yutao Sun, Zhifang Sui, and Furu Wei. 2025. Reinforcement pre-training. arXiv preprint arXiv:2506.08007. — [arXiv:2506.08007](https://arxiv.org/abs/2506.08007) [find](https://scholar.google.com/scholar?q=Qingxiu+Dong%2C+Li+Dong%2C+Yao+Tang%2C+Tianzhu+Ye%2C+Yutao+Sun%2C+Zhifang+Sui%2C+and+Furu+Wei%2E+2025%2E+Reinforcement+pre%2Dtraining%2E+arXiv+preprint+arXiv%3A2506%2E08007%2E+%E2%80%94+%5BarXiv%3A2506%2E08007%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2506%2E08)
20. Adam Tauman Kalai, Ofir Nachum, Santosh S. Vempala, and Edwin Zhang. 2025. Why language models hallucinate. Preprint, arXiv:2509.04664. — [arXiv:2509.04664](https://arxiv.org/abs/2509.04664) [find](https://scholar.google.com/scholar?q=Adam+Tauman+Kalai%2C+Ofir+Nachum%2C+Santosh+S%2E+Vempala%2C+and+Edwin+Zhang%2E+2025%2E+Why+language+models+hallucinate%2E+Preprint%2C+arXiv%3A2509%2E04664%2E+%E2%80%94+%5BarXiv%3A2509%2E04664%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2509%2E04664%29)
21. Ehsan Kamalloo, Aref Jafari, Xinyu Zhang, Nandan Thakur, and Jimmy Lin. 2023. Hagrid: A human-llm collaborative dataset for generative information-seeking with attribution. Preprint, arXiv:2307.16883. — [arXiv:2307.16883](https://arxiv.org/abs/2307.16883) [find](https://scholar.google.com/scholar?q=Ehsan+Kamalloo%2C+Aref+Jafari%2C+Xinyu+Zhang%2C+Nandan+Thakur%2C+and+Jimmy+Lin%2E+2023%2E+Hagrid%3A+A+human%2Dllm+collaborative+dataset+for+generative+information%2Dseeking+with+attribution%2E+Preprint%2C+arXiv%3A2307%2E16883%2E)
22. Zixuan Ke, Yijia Shao, Haowei Lin, Tatsuya Konishi, Gyuhak Kim, and Bing Liu. 2023. Continual pre-training of language models. arXiv preprint arXiv:2302.03241. [find](https://scholar.google.com/scholar?q=Zixuan+Ke%2C+Yijia+Shao%2C+Haowei+Lin%2C+Tatsuya+Konishi%2C+Gyuhak+Kim%2C+and+Bing+Liu%2E+2023%2E+Continual+pre%2Dtraining+of+language+models%2E+arXiv+preprint+arXiv%3A2302%2E03241%2E)
23. Nathan Lambert, Jacob Morrison, Valentina Pyatkin, Shengyi Huang, Hamish Ivison, Faeze Brahman, Lester James V Miranda, Alisa Liu, Nouha Dziri, Shane Lyu, and 1 others. 2024. [Tulu 3: Pushing frontiers in open language model post-training](https://md.research.elamur.ai/free/en/papers/lambert-2024-tulu-3-pushing-frontiers-open/). arXiv preprint arXiv:2411.15124.
24. Stephanie Lin, Jacob Hilton, and Owain Evans. 2022. Teaching models to express their uncertainty in words. arXiv preprint arXiv:2205.14334. [find](https://scholar.google.com/scholar?q=Stephanie+Lin%2C+Jacob+Hilton%2C+and+Owain+Evans%2E+2022%2E+Teaching+models+to+express+their+uncertainty+in+words%2E+arXiv+preprint+arXiv%3A2205%2E14334%2E)
25. Langming Liu, Haibin Chen, Yuhao Wang, Yujin Yuan, Shilei Liu, Wenbo Su, Xiangyu Zhao, and Bo Zheng. 2025. Eckgbench: Benchmarking large language models in e-commerce leveraging knowledge graph. In Proceedings of the 34th ACM International Conference on Information and Knowledge Management, pages 64… [find](https://scholar.google.com/scholar?q=Langming+Liu%2C+Haibin+Chen%2C+Yuhao+Wang%2C+Yujin+Yuan%2C+Shilei+Liu%2C+Wenbo+Su%2C+Xiangyu+Zhao%2C+and+Bo+Zheng%2E+2025%2E+Eckgbench%3A+Benchmarking+large+language+models+in+e%2Dcommerce+leveraging+knowledge+graph%2E+In+Pr)
26. Kangtao Lv, Haibin Chen, Yujin Yuan, Langming Liu, Shilei Liu, Yongwei Wang, Wenbo Su, and Bo Zheng. 2025. How to inject knowledge efficiently? knowledge infusion scaling law for pre-training large language models. Preprint, arXiv:2509.19371. [find](https://scholar.google.com/scholar?q=Kangtao+Lv%2C+Haibin+Chen%2C+Yujin+Yuan%2C+Langming+Liu%2C+Shilei+Liu%2C+Yongwei+Wang%2C+Wenbo+Su%2C+and+Bo+Zheng%2E+2025%2E+How+to+inject+knowledge+efficiently%3F+knowledge+infusion+scaling+law+for+pre%2Dtraining+large+la)
27. Alex Mallen, Akari Asai, Victor Zhong, Rajarshi Das, Daniel Khashabi, and Hannaneh Hajishirzi. 2023. When not to trust language models: Investigating effectiveness of parametric and non-parametric memories. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (V… [find](https://scholar.google.com/scholar?q=Alex+Mallen%2C+Akari+Asai%2C+Victor+Zhong%2C+Rajarshi+Das%2C+Daniel+Khashabi%2C+and+Hannaneh+Hajishirzi%2E+2023%2E+When+not+to+trust+language+models%3A+Investigating+effectiveness+of+parametric+and+non%2Dparametric+mem)
28. Potsawee Manakul, Adian Liusie, and Mark JF Gales. 2023. Selfcheckgpt: Zero-resource black-box hallucination detection for generative large language models. arXiv preprint arXiv:2303.08896. [find](https://scholar.google.com/scholar?q=Potsawee+Manakul%2C+Adian+Liusie%2C+and+Mark+JF+Gales%2E+2023%2E+Selfcheckgpt%3A+Zero%2Dresource+black%2Dbox+hallucination+detection+for+generative+large+language+models%2E+arXiv+preprint+arXiv%3A2303%2E08896%2E)
29. Kevin Meng, David Bau, Alex Andonian, and Yonatan Belinkov. 2022a. Locating and editing factual associations in gpt. Advances in neural information processing systems, 35:17359–17372. [find](https://scholar.google.com/scholar?q=Kevin+Meng%2C+David+Bau%2C+Alex+Andonian%2C+and+Yonatan+Belinkov%2E+2022a%2E+Locating+and+editing+factual+associations+in+gpt%2E+Advances+in+neural+information+processing+systems%2C+35%3A17359%E2%80%9317372%2E)
30. Kevin Meng, Arnab Sen Sharma, Alex Andonian, Yonatan Belinkov, and David Bau. 2022b. Mass-editing memory in a transformer. arXiv preprint arXiv:2210.07229. [find](https://scholar.google.com/scholar?q=Kevin+Meng%2C+Arnab+Sen+Sharma%2C+Alex+Andonian%2C+Yonatan+Belinkov%2C+and+David+Bau%2E+2022b%2E+Mass%2Dediting+memory+in+a+transformer%2E+arXiv+preprint+arXiv%3A2210%2E07229%2E)
31. Eric Mitchell, Charles Lin, Antoine Bosselut, Chelsea Finn, and Christopher D Manning. 2021. Fast model editing at scale. arXiv preprint arXiv:2110.11309. [find](https://scholar.google.com/scholar?q=Eric+Mitchell%2C+Charles+Lin%2C+Antoine+Bosselut%2C+Chelsea+Finn%2C+and+Christopher+D+Manning%2E+2021%2E+Fast+model+editing+at+scale%2E+arXiv+preprint+arXiv%3A2110%2E11309%2E)
32. OpenAI. 2024. [Gpt-4 technical report](https://md.research.elamur.ai/free/en/papers/openai-2023-gpt-4-technical-report/). arXiv:2303.08774.
33. Preprint, Ankit Pal, Logesh Kumar Umapathi, and Malaikannan Sankarasubbu. 2023. [Med-halt: Medical domain hallucination test for large language models](https://md.research.elamur.ai/free/en/papers/pal-2023-med-halt-medical-domain-hallucination/). Preprint, arXiv:2307.15343.
34. Richard Yuanzhe Pang, Weizhe Yuan, He He, Kyunghyun Cho, Sainbayar Sukhbaatar, and Jason E Weston. 2024. [Iterative reasoning preference optimization](https://md.research.elamur.ai/free/en/papers/pang-2024-iterative-reasoning-preference-optimization/). In The Thirty-eighth Annual Conference on Neural Information Processing Systems.
35. Alec Radford, Karthik Narasimhan, Tim Salimans, Ilya Sutskever, and 1 others. 2018. Improving language understanding by generative pre-training. [find](https://scholar.google.com/scholar?q=Alec+Radford%2C+Karthik+Narasimhan%2C+Tim+Salimans%2C+Ilya+Sutskever%2C+and+1+others%2E+2018%2E+Improving+language+understanding+by+generative+pre%2Dtraining%2E)
36. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. 2023. [Direct preference optimization: Your language model is secretly a reward model](https://md.research.elamur.ai/free/en/papers/rafailov-2023-direct-preference-optimization-language-model/). In Thirty-seventh Conference on Neural Information Processing Systems.
37. Adam Roberts, Colin Raffel, and Noam Shazeer. 2020. How much knowledge can you pack into the parameters of a language model? arXiv preprint arXiv:2002.08910. [find](https://scholar.google.com/scholar?q=Adam+Roberts%2C+Colin+Raffel%2C+and+Noam+Shazeer%2E+2020%2E+How+much+knowledge+can+you+pack+into+the+parameters+of+a+language+model%3F+arXiv+preprint+arXiv%3A2002%2E08910%2E)
38. Pranab Sahoo, Prabhash Meharia, Akash Ghosh, Sriparna Saha, Vinija Jain, and Aman Chadha. 2024. A comprehensive survey of hallucination in large language, image, video and audio foundation models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 11709–11724, Miami, Flo… [find](https://scholar.google.com/scholar?q=Pranab+Sahoo%2C+Prabhash+Meharia%2C+Akash+Ghosh%2C+Sriparna+Saha%2C+Vinija+Jain%2C+and+Aman+Chadha%2E+2024%2E+A+comprehensive+survey+of+hallucination+in+large+language%2C+image%2C+video+and+audio+foundation+models%2E+In+)
39. Christopher Sciavolino, Zexuan Zhong, Jinhyuk Lee, and Danqi Chen. 2022. [Simple entity-centric questions challenge dense retrievers](https://md.research.elamur.ai/free/en/papers/sciavolino-2021-simple-entity-centric-questions-challenge/). Preprint, arXiv:2109.08535.
40. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, YK Li, Yang Wu, and 1 others. 2024. [Deepseekmath: Pushing the limits of mathematical reasoning in open language models](https://md.research.elamur.ai/free/en/papers/shao-2024-deepseekmath-pushing-limits-mathematical-reasoning/). arXiv preprint arXiv:2402.03300.
41. Haizhou Shi, Zihao Xu, Hengyi Wang, Weiyi Qin, Wenyuan Wang, Yibin Wang, Zifeng Wang, Sayna Ebrahimi, and Hao Wang. 2024. Continual learning of large language models: A comprehensive survey. ACM Computing Surveys. [find](https://scholar.google.com/scholar?q=Haizhou+Shi%2C+Zihao+Xu%2C+Hengyi+Wang%2C+Weiyi+Qin%2C+Wenyuan+Wang%2C+Yibin+Wang%2C+Zifeng+Wang%2C+Sayna+Ebrahimi%2C+and+Hao+Wang%2E+2024%2E+Continual+learning+of+large+language+models%3A+A+comprehensive+survey%2E+ACM+Compu)
42. Linxin Song, Taiwei Shi, and Jieyu Zhao. 2025. The hallucination tax of reinforcement finetuning. Preprint, arXiv:2505.13988. [find](https://scholar.google.com/scholar?q=Linxin+Song%2C+Taiwei+Shi%2C+and+Jieyu+Zhao%2E+2025%2E+The+hallucination+tax+of+reinforcement+finetuning%2E+Preprint%2C+arXiv%3A2505%2E13988%2E)
43. Kai Sun, Yifan Ethan Xu, Hanwen Zha, Yue Liu, and Xin Luna Dong. 2024. Head-to-tail: How knowledgeable are large language models (llms)? a.k.a. will llms replace knowledge graphs? Preprint, arXiv:2308.10168. [find](https://scholar.google.com/scholar?q=Kai+Sun%2C+Yifan+Ethan+Xu%2C+Hanwen+Zha%2C+Yue+Liu%2C+and+Xin+Luna+Dong%2E+2024%2E+Head%2Dto%2Dtail%3A+How+knowledgeable+are+large+language+models+%28llms%29%3F+a%2Ek%2Ea%2E+will+llms+replace+knowledge+graphs%3F+Preprint%2C+arXiv%3A2308)
44. Mirac Suzgun, Nathan Scales, Nathanael Schärli, Sebastian Gehrmann, Yi Tay, Hyung Won Chung, Aakanksha Chowdhery, Quoc V Le, Ed H Chi, Denny Zhou, , and Jason Wei. 2022. [Challenging big-bench tasks and whether chain-of-thought can solve them](https://md.research.elamur.ai/free/en/papers/suzgun-2023-challenging-big-bench-tasks-whether/). arXiv preprint arXiv:2210.09261.
45. SMTI Tonmoy, SM Zaman, Vinija Jain, Anku Rani, Vipula Rawte, Aman Chadha, and Amitava Das. 2024. [A comprehensive survey of hallucination mitigation techniques in large language models](https://md.research.elamur.ai/free/en/papers/tonmoy-2024-comprehensive-survey-hallucination-mitigation-techniques/). arXiv preprint arXiv:2401.01313, 6.
46. Yulong Chen, and 1 others. 2025. [Siren’s song in the ai ocean: A survey on hallucination in large language models](https://md.research.elamur.ai/free/en/papers/zhang-2023-siren-s-song-ai-ocean/). Computational Linguistics, pages 1–46.
47. Ce Zheng, Lei Li, Qingxiu Dong, Yuxuan Fan, Zhiyong Wu, Jingjing Xu, and Baobao Chang. 2023. Can we edit factual knowledge by in-context learning? arXiv preprint arXiv:2305.12740. [find](https://scholar.google.com/scholar?q=Ce+Zheng%2C+Lei+Li%2C+Qingxiu+Dong%2C+Yuxuan+Fan%2C+Zhiyong+Wu%2C+Jingjing+Xu%2C+and+Baobao+Chang%2E+2023%2E+Can+we+edit+factual+knowledge+by+in%2Dcontext+learning%3F+arXiv+preprint+arXiv%3A2305%2E12740%2E)
48. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, and 1 others. 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.
49. Liyuan Wang, Xingxing Zhang, Hang Su, and Jun Zhu. 2024a. A comprehensive survey of continual learning: Theory, method and application. IEEE transactions on pattern analysis and machine intelligence, 46(8):5362–5383. [find](https://scholar.google.com/scholar?q=Liyuan+Wang%2C+Xingxing+Zhang%2C+Hang+Su%2C+and+Jun+Zhu%2E+2024a%2E+A+comprehensive+survey+of+continual+learning%3A+Theory%2C+method+and+application%2E+IEEE+transactions+on+pattern+analysis+and+machine+intelligence%2C+)
50. Peiyi Wang, Lei Li, Zhihong Shao, Runxin Xu, Damai Dai, Yifei Li, Deli Chen, Yu Wu, and Zhifang Sui. 2024b. [Math-shepherd: Verify and reinforce LLMs step-by-step without human annotations](https://md.research.elamur.ai/free/en/papers/wang-2024-math-shepherd-verify-reinforce-llms/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Pape…
51. Song Wang, Yaochen Zhu, Haochen Liu, Zaiyi Zheng, Chen Chen, and Jundong Li. 2024c. Knowledge editing for [large language models: A survey](https://md.research.elamur.ai/free/en/papers/minaee-2024-large-language-models-survey/). ACM Computing Surveys, 57(3):1–37.
52. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc Le, and Denny Zhou. 2023. [Chain-of-thought prompting elicits reasoning in large language models](https://md.research.elamur.ai/free/en/papers/wei-2022-chain-thought-prompting-elicits-reasoning/). Preprint, arXiv:2201.11903.
53. Taiqiang Wu, Runming Yang, Jiayi Li, Pengfei Hu, Yik-Chung Wu, Ngai Wong, and Yujiu Yang. 2025. Shadow-ft: Tuning instruct model via training on paired base model. Preprint, arXiv:2505.12716. — [arXiv:2505.12716](https://arxiv.org/abs/2505.12716) [find](https://scholar.google.com/scholar?q=Taiqiang+Wu%2C+Runming+Yang%2C+Jiayi+Li%2C+Pengfei+Hu%2C+Yik%2DChung+Wu%2C+Ngai+Wong%2C+and+Yujiu+Yang%2E+2025%2E+Shadow%2Dft%3A+Tuning+instruct+model+via+training+on+paired+base+model%2E+Preprint%2C+arXiv%3A2505%2E12716%2E+%E2%80%94+%5BarXiv)
54. Ziwei Xu, Sanjay Jain, and Mohan Kankanhalli. 2024. [Hallucination is inevitable: An innate limitation of large language models](https://md.research.elamur.ai/free/en/papers/xu-2024-hallucination-inevitable-innate-limitation-large/). arXiv preprint arXiv:2401.11817.
55. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, Chujie Zheng, Dayiheng Liu, Fan Zhou, Fei Huang, Feng Hu, Hao Ge, Haoran Wei, Huan Lin, Jialong Tang, and 41 others. 2025. [Qwen3 technical report](https://md.research.elamur.ai/free/en/papers/yang-2025-qwen3-technical-report/). Preprint, arXiv:2505.09388.
56. Yue Zhang, Yafu Li, Leyang Cui, Deng Cai, Lemao Liu, Tingchen Fu, Xinting Huang, Enbo Zhao, Yu Zhang, [find](https://scholar.google.com/scholar?q=Yue+Zhang%2C+Yafu+Li%2C+Leyang+Cui%2C+Deng+Cai%2C+Lemao+Liu%2C+Tingchen+Fu%2C+Xinting+Huang%2C+Enbo+Zhao%2C+Yu+Zhang%2C)

*Updated 1 November 2026*