Elamur Research⌕
☰
home / papers / 2025 / LoGU: Long-form Generation with Uncertainty Expressions

LoGU: Long-form Generation with Uncertainty Expressions

Paper · 2025 · Ruihan Yang, Caiqi Zhang, Zhisong Zhang et al.

Summary

Work on getting LLMs to admit what they do not know has mostly targeted short-form QA, where one confidence signal covers a single answer. Long responses mix true and false claims, so uncertainty must be expressed claim by claim. The authors define this as the LoGU task (Long-form Generation with Uncertainty). They identify two failure modes: Uncertainty Suppression, where models avoid hedging, and Uncertainty Misalignment, where hedges land on the wrong facts.

To build training data, model outputs are split into atomic claims and fact-checked with FActScore against retrieved evidence. An auxiliary LLM (GPT-4) then rewrites selected claims as uncertainty statements and merges everything back into fluent text. Training happens in two stages. LoGU-SFT teaches the model to hedge at all. LoGU-DPO then uses preference pairs to steer hedging toward facts the model actually gets wrong.

On three long-form benchmarks, the method raises factual accuracy and reduces false statements while keeping responses comprehensive. Its hedges are also more specific than the vague caveats produced by baselines. This makes long answers more trustworthy and their gaps easier to see.

Key points

  • Two metrics are defined:
    • Factual Accuracy (FA) is the share of supported claims among all definite claims.
    • Uncertain Accuracy (UA) is the share of uncertainty expressions that target genuinely unknown, fine-grained details.
  • Revision uses two operators:
    • The positive operator (Γ_pos) turns refuted claims into hedges.
    • The negative operator (Γ_neg) wrongly hedges supported claims instead.
    • Rewriting draws on ten hand-crafted uncertainty patterns, with down-sampling to limit how often hedges appear.
  • LoGU-DPO prefers the positive revision R_pos over both the original response and R_neg, and prefers the original over R_neg.
  • Evaluation covers three datasets without gold answers: Bios, LongFact and WildHallu. For this reason, ROUGE-L and recall are not reported.
  • In human evaluation, outputs with about 20% hedged content were judged most helpful, with only slightly lower fluency.
  • Uncertainty is expressed as binary rather than graded. Code is available at github.com/rhyang2021/LoGU.

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

🎓University of CambridgeCambridge · United Kingdom243 works · 121,396 citations🎓Fudan UniversityShanghai · China167 works · 20,987 citations🎓Chinese University of Hong KongHong Kong · China126 works · 14,073 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

Cite

Cite as Ruihan Yang, Caiqi Zhang, Zhisong Zhang, Xinting Huang, Sen Yang, Nigel Collier, Dong Yu and Deqing Yang. 2025. LoGU: Long-form Generation with Uncertainty Expressions. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2025.acl-long.928

@inproceedings{yang2025logu,
  title = {{LoGU: Long-form Generation with Uncertainty Expressions}},
  author = {Ruihan Yang, Caiqi Zhang, Zhisong Zhang, Xinting Huang, Sen Yang, Nigel Collier, Dong Yu and Deqing Yang},
  booktitle = {Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2025},
  doi = {10.18653/v1/2025.acl-long.928},
  url = {https://doi.org/10.18653/v1/2025.acl-long.928}
}

References

42 references, 16 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report. ArXiv preprint, abs/2303.08774.
  2. Yushi Bai, Xin Lv, Jiajie Zhang, Hongchang Lyu, Jiankai Tang, Zhidian Huang, Zhengxiao Du, Xiao Liu, Aohan Zeng, Lei Hou, et al. 2023. Longbench: A bilingual, multitask benchmark for long context understanding. ArXiv preprint, abs/2308.14508.
  3. Neil Band, Xuechen Li, Tengyu Ma, and Tatsunori Hashimoto. 2024. Linguistic calibration of long-form generations. In Forty-first International Conference on Machine Learning, ICML 2024, Vienna, Austria, July 21-27, 2024. OpenReview.net. find
  4. Yupeng Chang, Xu Wang, Jindong Wang, Yuan Wu, Linyi Yang, Kaijie Zhu, Hao Chen, Xiaoyuan Yi, Cunxiang Wang, Yidong Wang, et al. 2024. A survey on evaluation of large language models. ACM Transactions on Intelligent Systems and Technology, 15(3):1–45.
  5. Yulong Chen, Yang Liu, Jianhao Yan, Xuefeng Bai, Ming Zhong, Yinghao Yang, Ziyi Yang, Chenguang Zhu, and Yue Zhang. 2024. See what LLMs cannot answer: A self-challenge framework for uncovering LLM weaknesses. In First Conference on Language Modeling. find
  6. Nicola De Cao, Wilker Aziz, and Ivan Titov. 2021. Editing factual knowledge in language models. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 6491– 6506, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics. find
  7. Ekaterina Fadeeva, Roman Vashurin, Akim Tsvigun, Artem Vazhentsev, Sergey Petrakov, Kirill Fedyanin, Daniil Vasilev, Elizaveta Goncharova, Alexander Panchenko, Maxim Panov, Timothy Baldwin, and Artem Shelmanov. 2023. LM-polygraph: Uncertainty estimation for language models. In Proceedings of the 202… find
  8. Sebastian Farquhar, Jannik Kossen, Lorenz Kuhn, and Yarin Gal. 2024. Detecting hallucinations in large language models using semantic entropy. Nature, 630(8017):625–630. find
  9. 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? ArXiv preprint, abs/2405.05904. find
  10. Muhammad Usman Hadi, Rizwan Qureshi, Abbas Shah, Muhammad Irfan, Anas Zafar, Muhammad Bilal Shaikh, Naveed Akhtar, Jia Wu, Seyedali Mirjalili, et al. 2023. A survey on large language models: Applications, challenges, limitations, and practical usage. Authorea Preprints. find
  11. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. Lora: Low-rank adaptation of large language models. In The Tenth International Conference on Learning Representations, ICLR 2022, Virtual Event, April 25-29, 2022. OpenReview.net.
  12. Association for Computational Linguistics: NAACL 2025, pages 1272–1287, Albuquerque, New Mexico. Association for Computational Linguistics. find
  13. Saurav Kadavath, Tom Conerly, Amanda Askell, Tom Henighan, Dawn Drain, Ethan Perez, Nicholas Schiefer, Zac Hatfield-Dodds, Nova DasSarma, Eli Tran-Johnson, Scott Johnston, Sheer El-Showk, Andy Jones, Nelson Elhage, Tristan Hume, Anna Chen, Yuntao Bai, Sam Bowman, Stanislav Fort, Deep Ganguli, Danny… find
  14. Katie Kang, Eric Wallace, Claire Tomlin, Aviral Kumar, and Sergey Levine. 2024. Unfamiliar finetuning examples control how language models hallucinate. ArXiv preprint, abs/2403.05612.
  15. Diederik P. Kingma and Jimmy Ba. 2015. Adam: A method for stochastic optimization. In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings.
  16. Lorenz Kuhn, Yarin Gal, and Sebastian Farquhar. 2023. Semantic uncertainty: Linguistic invariances for uncertainty estimation in natural language generation. In The Eleventh International Conference on Learning Representations, ICLR 2023, Kigali, Rwanda, May 1-5, 2023. OpenReview.net. find
  17. Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph Gonzalez, Hao Zhang, and Ion Stoica. 2023. Efficient memory management for large language model serving with pagedattention. In Proceedings of the 29th Symposium on Operating Systems Principles, pages 611–626.
  18. J Richard Landis and Gary G Koch. 1977. The measurement of observer agreement for categorical data. biometrics, pages 159–174. find
  19. Yukun Huang, Yixin Liu, Raghuveer Thirukovalluru, Arman Cohan, and Bhuwan Dhingra. 2024. Calibrating long-form generations from large language models. Preprint, arXiv:2402.06544. find
  20. Jiaqi Li, Yixuan Tang, and Yi Yang. 2024. Know the unknown: An uncertainty-sensitive method for llm instruction tuning. find
  21. Albert Q. Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, Lélio Renard Lavaud, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix,… find
  22. Sheng-Chieh Lin, Luyu Gao, Barlas Oguz, Wenhan Xiong, Jimmy Lin, Scott Yih, and Xilun Chen. 2024. FLAME : Factuality-aware alignment for large language models. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, NeurIPS 2024, Vanc… find
  23. Lekang Jiang, Caiqi Zhang, Pascal A. Scherz, and Stefan Goetz. 2025. Can large language models generate high-quality patent claims? In Findings of the Stephanie Lin, Jacob Hilton, and Owain Evans. 2022. Teaching models to express their uncertainty in words. ArXiv preprint, abs/2205.14334. find
  24. Aman Madaan, Niket Tandon, Prakhar Gupta, Skyler Hallinan, Luyu Gao, Sarah Wiegreffe, Uri Alon, Nouha Dziri, Shrimai Prabhumoye, Yiming Yang, Shashank Gupta, Bodhisattwa Prasad Majumder, Katherine Hermann, Sean Welleck, Amir Yazdanbakhsh, and Peter Clark. 2023. Self-refine: Iterative refinement with… find
  25. Meta. 2024. Llama 3 model card. find
  26. Sewon Min, Kalpesh Krishna, Xinxi Lyu, Mike Lewis, Wen-tau Yih, Pang Koh, Mohit Iyyer, Luke Zettlemoyer, and Hannaneh Hajishirzi. 2023. FActScore: Fine-grained atomic evaluation of factual precision in long form text generation. In Proceedings of the 2023 Conference on Empirical Methods in Natural L…
  27. 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. In Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Sy…
  28. Ivan Stelmakh, Yi Luan, Bhuwan Dhingra, and Ming-Wei Chang. 2022. ASQA: Factoid questions meet long-form answers. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 8273–8288, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
  29. Katherine Tian, Eric Mitchell, Huaxiu Yao, Christopher D. Manning, and Chelsea Finn. 2024. Fine-tuning language models for factuality. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024. OpenReview.net.
  30. Miao Xiong, Zhiyuan Hu, Xinyang Lu, Yifei Li, Jie Fu, Junxian He, and Bryan Hooi. 2024a. Can llms express their uncertainty? an empirical evaluation of confidence elicitation in llms. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024. Ope… find
  31. Miao Xiong, Zhiyuan Hu, Xinyang Lu, Yifei Li, Jie Fu, Junxian He, and Bryan Hooi. 2024b. Can llms express their uncertainty? an empirical evaluation of confidence elicitation in llms. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024. Ope… find
  32. Tianyang Xu, Shujin Wu, Shizhe Diao, Xiaoze Liu, Xingyao Wang, Yangyi Chen, and Jing Gao. 2024. Sayself: Teaching llms to express confidence with self-reflective rationales. Preprint, arXiv:2405.20974. find
  33. Ruihan Yang, Caiqi Zhang, Zhisong Zhang, Xinting Huang, Dong Yu, Nigel Collier, and Deqing Yang. 2025. Uncle: Uncertainty expressions in long-form generation. Preprint, arXiv:2505.16922. — [arXiv:2505.16922](https://arxiv.org/abs/2505.16922) find
  34. Tao Yuan, Xuefei Ning, Dong Zhou, Zhijie Yang, Shiyao Li, Minghui Zhuang, Zheyue Tan, Zhuyu Yao, Dahua Lin, Boxun Li, et al. 2024. Lv-eval: A balanced long-context benchmark with 5 length levels up to 256k. ArXiv preprint, abs/2402.05136.
  35. Caiqi Zhang, Zhijiang Guo, and Andreas Vlachos. 2024a. Do we need language-specific fact-checking models? the case of Chinese. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 1899–1914, Miami, Florida, USA. Association for Computational Linguistics. find
  36. Caiqi Zhang, Fangyu Liu, Marco Basaldella, and Nigel Collier. 2024b. Luq: Long-text uncertainty quantification for llms. Preprint, arXiv:2403.20279.
  37. Katherine Tian, Eric Mitchell, Allan Zhou, Archit Sharma, Rafael Rafailov, Huaxiu Yao, Chelsea Finn, and Christopher Manning. 2023. Just ask for calibration: Strategies for eliciting calibrated confidence scores from language models fine-tuned with human feedback. In Proceedings of the 2023 Conferen… find
  38. Hanning Zhang, Shizhe Diao, Yong Lin, Yi Fung, Qing Lian, Xingyao Wang, Yangyi Chen, Heng Ji, and Tong Zhang. 2024c. R-tuning: Instructing large language models to say ‘I don’t know’. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics… find
  39. Jerry Wei, Chengrun Yang, Xinying Song, Yifeng Lu, Nathan Hu, Jie Huang, Dustin Tran, Daiyi Peng, Ruibo Liu, Da Huang, Cosmo Du, and Quoc V. Le. 2024. Long-form factuality in large language models. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Proce…
  40. Xiaoying Zhang, Baolin Peng, Ye Tian, Jingyan Zhou, Lifeng Jin, Linfeng Song, Haitao Mi, and Helen Meng. 2024d. Self-alignment for factuality: Mitigating hallucinations in LLMs via self-evaluation. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1:…
  41. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, Yifan Du, Chen Yang, Yushuo Chen, Zhipeng Chen, Jinhao Jiang, Ruiyang Ren, Yifan Li, Xinyu Tang, Zikang Liu, Peiyu Liu, Jian-Yun Nie, and Ji-Rong Wen. 2023. A survey of l…
  42. Wenting Zhao, Tanya Goyal, Yu Ying Chiu, Liwei Jiang, Benjamin Newman, Abhilasha Ravichander, Khyathi Chandu, Ronan Le Bras, Claire Cardie, Yuntian Deng, and Yejin Choi. 2024. Wildhallucinations: Evaluating long-form factuality in llms with real-world entity queries. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/yang-2025-logu-long-form-generation-uncertainty/