Elamur Research⌕
☰
home / papers / 2025 / Removal of Hallucination on Hallucination: Debate-Augmented RAG

Removal of Hallucination on Hallucination: Debate-Augmented RAG

Paper · 2025 · Wentao Hu, Wengyu Zhang, Yiyang Jiang et al.

Summary

Retrieval-augmented generation is supposed to reduce LLM hallucinations, but the authors argue it can make them worse. Irrelevant or misleading passages, such as a film sharing a band's name, can steer the model toward a wrong answer. Even correct passages mixed with noise can do this. They call this compounding effect "Hallucination on Hallucination."

Their remedy is DRAG (Debate-Augmented RAG), a training-free framework that adds Multi-Agent Debate to both pipeline stages:

  • Retrieval stage: proponent, opponent and judge agents argue over whether the current query pool is sufficient. They iteratively revise it until the judge decides enough evidence has been gathered.
  • Generation stage: agents receive deliberately unequal access to the retrieved evidence and debate adversarially. This discourages blind trust in retrieved text.

Tested on six benchmarks covering three knowledge-intensive task types, DRAG is compared with Naive RAG, IRCoT, Iter-RetGen, FLARE, Self-RAG, SuRe and plain MAD. The authors report more reliable retrieval and better factual accuracy. The work matters because it treats RAG errors as a whole-pipeline problem rather than fixing a single stage.

Key points

  • DRAG has two debate stages: Retrieval Debate (f RetDebate), which refines a dynamic query pool, and Response Debate (f ResDebate), which uses asymmetric information roles. Code is available at github.com/Huenao/Debate-Augmented-RAG.
  • In ablations, each debate stage alone beats Naive RAG on most tasks. Giving all agents equal evidence access (the "- Info Asymmetry" variant) lowers results on every dataset.
  • On StrategyQA, DRAG averages 2.08 retriever calls and 10.36 LLM calls. Of these, 3.36 occur in retrieval and 7 in the response stage, a cost comparable to multi-turn methods like IRCoT.
  • The retrieval debate stops adaptively through a judge agent. The response debate uses a fixed number of rounds, which wastes calls on simple single-hop questions and can cause "problem drift."
  • Proposed future work includes confidence-based early stopping and agent pruning in the response stage.

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

🎓Hong Kong Polytechnic UniversityHong Kong · China84 works · 4,956 citations🎓Sichuan UniversityChengdu · China14 works · 1,293 citations

 Fields, topics and concepts

📚 Shelves 6The 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 worksComputer Vision and Pattern Recognition · 2,869 worksInformation Systems · 1,437 works

Cite

Cite as Wentao Hu, Wengyu Zhang, Yiyang Jiang, Chen Jason Zhang, Xiao-Yong Wei and Li Qing. 2025. Removal of Hallucination on Hallucination: Debate-Augmented RAG. 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.770

@inproceedings{hu2025removal,
  title = {{Removal of Hallucination on Hallucination: Debate-Augmented RAG}},
  author = {Wentao Hu, Wengyu Zhang, Yiyang Jiang, Chen Jason Zhang, Xiao-Yong Wei and Li Qing},
  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.770},
  url = {https://doi.org/10.18653/v1/2025.acl-long.770}
}

References

38 references, 27 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 arXiv:2303.08774.
  2. Akari Asai, Zeqiu Wu, Yizhong Wang, Avirup Sil, and Hannaneh Hajishirzi. 2023. Self-rag: Learning to retrieve, generate, and critique through self-reflection. arXiv preprint arXiv:2310.11511.
  3. Jonas Becker. 2024. Multi-agent large language models for conversational task-solving. arXiv preprint arXiv:2410.22932. — [arXiv:2410.22932](https://arxiv.org/abs/2410.22932) find
  4. Yilun Du, Shuang Li, Antonio Torralba, Joshua B. Tenenbaum, and Igor Mordatch. 2024. Improving factuality and reasoning in language models through multiagent debate. In Proceedings of the 41st International Conference on Machine Learning, ICML’24. JMLR.org.
  5. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. 2024. The llama 3 herd of models. arXiv preprint arXiv:2407.21783.
  6. Zhangyin Feng, Xiaocheng Feng, Dezhi Zhao, Maojin Yang, and Bing Qin. 2024. Retrieval-generation synergy augmented large language models. In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 11661–11665. IEEE. find
  7. Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yi Dai, Jiawei Sun, and Haofen Wang. 2023. Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997.
  8. Mor Geva, Daniel Khashabi, Elad Segal, Tushar Khot, Dan Roth, and Jonathan Berant. 2021. Did aristotle use a laptop? a question answering benchmark with implicit reasoning strategies. Transactions of the Association for Computational Linguistics, 9:346– 361. find
  9. Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat, and Mingwei Chang. 2020. Retrieval augmented language model pre-training. In International conference on machine learning, pages 3929–3938. PMLR.
  10. Xanh Ho, Anh-Khoa Duong Nguyen, Saku Sugawara, and Akiko Aizawa. 2020. Constructing a multi-hop QA dataset for comprehensive evaluation of reasoning steps. In Proceedings of the 28th International Conference on Computational Linguistics, pages 6609–6625, Barcelona, Spain (Online). International Comm…
  11. Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, et al. 2024. A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. ACM Transactions on Information Systems.
  12. Gautier Izacard and Edouard Grave. 2021. Leveraging passage retrieval with generative models for open domain question answering. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 874–880, Online. Association for Computa…
  13. Gautier Izacard, Patrick Lewis, Maria Lomeli, Lucas Hosseini, Fabio Petroni, Timo Schick, Jane Dwivedi-Yu, Armand Joulin, Sebastian Riedel, and Edouard Grave. 2022. Few-shot learning with retrieval augmented language models. arXiv preprint arXiv:2208.03299.
  14. 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. ACM Computing Surveys, 55(12):1–38.
  15. Zhengbao Jiang, Frank Xu, Luyu Gao, Zhiqing Sun, Qian Liu, Jane Dwivedi-Yu, Yiming Yang, Jamie Callan, and Graham Neubig. 2023. Active retrieval augmented generation. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 7969–7992, Singapore. Association fo…
  16. Jiajie Jin, Yutao Zhu, Xinyu Yang, Chenghao Zhang, and Zhicheng Dou. 2024. Flashrag: A modular toolkit for efficient retrieval-augmented generation research. CoRR, abs/2405.13576. find
  17. Mandar Joshi, Eunsol Choi, Daniel Weld, and Luke Zettlemoyer. 2017. TriviaQA: A large scale distantly supervised challenge dataset for reading comprehension. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1601–1611, Vancouver…
  18. Vladimir Karpukhin, Barlas Oguz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih. 2020. Dense passage retrieval for open-domain question answering. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 6769–6781, Onli…
  19. Jaehyung Kim, Jaehyun Nam, Sangwoo Mo, Jongjin Park, Sang-Woo Lee, Minjoon Seo, Jung-Woo Ha, and Jinwoo Shin. 2024. Sure: Summarizing retrievals using answer candidates for open-domain qa of llms. arXiv preprint arXiv:2404.13081. — [arXiv:2404.13081](https://arxiv.org/abs/2404.13081) find
  20. Tom Kwiatkowski, Jennimaria Palomaki, Olivia Redfield, Michael Collins, Ankur Parikh, Chris Alberti, Danielle Epstein, Illia Polosukhin, Jacob Devlin, Kenton Lee, Kristina Toutanova, Llion Jones, Matthew Kelcey, Ming-Wei Chang, Andrew M. Dai, Jakob Uszkoreit, Quoc Le, and Slav Petrov. 2019. Natural…
  21. Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, et al. 2020. Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems, 33:9459–9474.
  22. Tian Liang, Zhiwei He, Wenxiang Jiao, Xing Wang, Yan Wang, Rui Wang, Yujiu Yang, Shuming Shi, and Zhaopeng Tu. 2024. Encouraging divergent thinking in large language models through multi-agent debate. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 17…
  23. 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
  24. Keheng Wang, Feiyu Duan, Peiguang Li, Sirui Wang, and Xunliang Cai. 2025. LLMs know what they need: Leveraging a missing information guided framework to empower retrieval-augmented generation. In Proceedings of the 31st International Conference on Computational Linguistics, pages 2379–2400, Abu Dhab…
  25. Chong-Wah Ngo, Yu-Gang Jiang, Xiaoyong Wei, Feng Wang, Wanlei Zhao, Hung-Khoon Tan, and Xiao Wu. 2007. Experimenting vireo-374: Bag-of-visual-words and visual-based ontology for semantic video indexing and search. find
  26. Liang Wang, Nan Yang, Xiaolong Huang, Binxing Jiao, Linjun Yang, Daxin Jiang, Rangan Majumder, and Furu Wei. 2022. Text embeddings by weakly-supervised contrastive pre-training. arXiv preprint arXiv:2212.03533.
  27. Timo Schick, Jane Dwivedi-Yu, Roberto Dessì, Roberta Raileanu, Maria Lomeli, Eric Hambro, Luke Zettlemoyer, Nicola Cancedda, and Thomas Scialom. 2023. Toolformer: Language models can teach themselves to use tools. Advances in Neural Information Processing Systems, 36:68539–68551.
  28. Xiao-Yong Wei and Chong-Wah Ngo. 2008. Fusing semantics, observability, reliability and diversity of concept detectors for video search. In Proceedings of the 16th ACM international conference on Multimedia, pages 81–90. find
  29. Zhihong Shao, Yeyun Gong, Yelong Shen, Minlie Huang, Nan Duan, and Weizhu Chen. 2023. Enhancing retrieval-augmented large language models with iterative retrieval-generation synergy. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 9248–9274, Singapore. Association for…
  30. Weijia Shi, Sewon Min, Michihiro Yasunaga, Minjoon Seo, Richard James, Mike Lewis, Luke Zettlemoyer, and Wen-tau Yih. 2024. REPLUG: Retrieval-augmented black-box language models. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Hum…
  31. Weihang Su, Yichen Tang, Qingyao Ai, Zhijing Wu, and Yiqun Liu. 2024. DRAGIN: Dynamic retrieval augmented generation based on the real-time information needs of large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers),…
  32. 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. arXiv preprint arXiv:2307.09288.
  33. Fangyuan Xu, Weijia Shi, and Eunsol Choi. 2023. Recomp: Improving retrieval-augmented lms with compression and selective augmentation. arXiv preprint arXiv:2310.04408.
  34. Zhilin Yang, Peng Qi, Saizheng Zhang, Yoshua Bengio, William Cohen, Ruslan Salakhutdinov, and Christopher D. Manning. 2018. HotpotQA: A dataset for diverse, explainable multi-hop question answering. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 2369…
  35. Zhangyue Yin, Qiushi Sun, Cheng Chang, Qipeng Guo, Junqi Dai, Xuanjing Huang, and Xipeng Qiu. 2023. Exchange-of-thought: Enhancing large language model capabilities through cross-model communication. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 151… find
  36. Tian Yu, Shaolei Zhang, and Yang Feng. 2024. Auto-rag: Autonomous retrieval-augmented generation for large language models. arXiv preprint arXiv:2411.19443. — [arXiv:2411.19443](https://arxiv.org/abs/2411.19443) find
  37. Changmeng Zheng, Dayong Liang, Wengyu Zhang, Xiao-Yong Wei, Tat-Seng Chua, and Qing Li. 2024. A picture is worth a graph: A blueprint debate paradigm for multimodal reasoning. In Proceedings of the 32nd ACM International Conference on Multimedia, pages 419–428. find
  38. Harsh Trivedi, Niranjan Balasubramanian, Tushar Khot, and Ashish Sabharwal. 2023. Interleaving retrieval with chain-of-thought reasoning for knowledge-intensive multi-step questions. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), p…

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/hu-2025-removal-hallucination-hallucination-debate-augmented/