Elamur Research⌕
☰
home / papers / 2025 / FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation

FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation

Paper · 2025 · Qinggang Zhang, Zhishang Xiang, Yilin Xiao et al.

Summary

Retrieval-augmented LLMs often fail to stay faithful to retrieved passages when those passages contradict what the model learned during pre-training. The authors show that current faithfulness techniques, both prompt-based and decoding-based (e.g., CAD, COIECD), mostly push the model to discard its own knowledge. This reduces over-confidence errors but makes the model more likely to adopt or misread misleading context. FaithfulRAG handles the conflict explicitly in three steps:

  • Self-fact mining: the LLM writes out what it believes about the question as fine-grained facts.
  • Contextual knowledge alignment: these facts are compared with the retrieved text to locate the conflicting segments.
  • Self-thinking: the model reasons over those conflicts before answering.

The approach requires no fine-tuning. Evaluated on FaithEval, MuSiQue, SQuAD and RealtimeQA, it outperforms state-of-the-art faithful RAG baselines. The work suggests that faithfulness improves when parametric knowledge is reconciled with the context rather than suppressed. This is relevant to any RAG system built over updated or domain-specific documents.

Key points

  • In a preliminary study, Llama3.1-8b-instruct, Qwen2.5-7b-instruct and Mistral-7b-instruct lost between 9.7% and 29.9% in accuracy on MuSiQue and SQuAD once knowledge conflicts were introduced.
  • On average, existing faithful methods cut unfaithful (over-confidence) errors by 6.65% but increased incorrect-match errors by 6.42%. With decoding-based methods, incorrect-match errors reached 14.3% on MuSiQue and 16.4% on SQuAD.
  • The framework is training-free and targets only the generation stage of RAG, not retrieval.
  • The four public benchmarks are FaithEval, whose contexts are counterfactual, inconsistent or unanswerable; the KRE versions of MuSiQue and SQuAD; and RealtimeQA.
  • The method is limited to text and does not handle multimodal input. Code is available at github.com/DeepLearnXMU/Faithful-RAG.

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, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksInformation Systems · 1,730 works

🧭 Topics 1Research subjects and languages the work is about.Information retrieval & search · 552 works

Cite

Cite as Qinggang Zhang, Zhishang Xiang, Yilin Xiao, Le Wang, Junhui Li, Xinrun Wang and Jinsong Su. 2025. FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation. 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.1062

@inproceedings{zhang2025faithfulrag,
  title = {{FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation}},
  author = {Qinggang Zhang, Zhishang Xiang, Yilin Xiao, Le Wang, Junhui Li, Xinrun Wang and Jinsong Su},
  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.1062},
  url = {https://doi.org/10.18653/v1/2025.acl-long.1062}
}

References

43 references, 23 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. AI Anthropic. 2024. The claude 3 model family: Opus, sonnet, haiku. Claude-3 Model Card. find
  2. Akari Asai, Zeqiu Wu, Yizhong Wang, Avirup Sil, and Hannaneh Hajishirzi. 2023. Self-rag: Learning to retrieve, generate, and critique through self-reflection. Preprint, arXiv:2310.11511.
  3. Baolong Bi, Shaohan Huang, Yiwei Wang, Tianchi Yang, Zihan Zhang, Haizhen Huang, Lingrui Mei, Junfeng Fang, Zehao Li, Furu Wei, et al. 2024a. Context-dpo: Aligning language models for context-faithfulness. arXiv preprint arXiv:2412.15280. find
  4. Baolong Bi, Shenghua Liu, Yiwei Wang, Lingrui Mei, Junfeng Fang, Hongcheng Gao, Shiyu Ni, and Xueqi Cheng. 2024b. Is factuality enhancement a free lunch for llms? better factuality can lead to worse context-faithfulness. arXiv preprint arXiv:2404.00216. — [arXiv:2404.00216](https://arxiv.org/abs/2404.00216) find
  5. Zhiwei Cao, Qian Cao, Yu Lu, Ningxin Peng, Luyang Huang, Shanbo Cheng, and Jinsong Su. 2024. Retaining key information under high compression ratios: Query-guided compressor for llms. Preprint, arXiv:2406.02376. find
  6. Shengyuan Chen, Qinggang Zhang, Junnan Dong, Wen Hua, Qing Li, and Xiao Huang. 2024. Entity alignment with noisy annotations from large language models. arXiv preprint arXiv:2405.16806. — [arXiv:2405.16806](https://arxiv.org/abs/2405.16806) find
  7. Junfeng Fang, Houcheng Jiang, Kun Wang, Yunshan Ma, Shi Jie, Xiang Wang, Xiangnan He, and Tat-Seng Chua. 2025. Alphaedit: Null-space constrained knowledge editing for language models. ICLR. find
  8. 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.
  9. Zijin Hong, Zheng Yuan, Hao Chen, Qinggang Zhang, Feiran Huang, and Xiao Huang. 2024. Knowledge-to-sql: Enhancing sql generation with data expert llm. arXiv preprint arXiv:2402.11517. find
  10. Cheng-Yu Hsieh, Yung-Sung Chuang, Chun-Liang Li, Zifeng Wang, Long T Le, Abhishek Kumar, James Glass, Alexander Ratner, Chen-Yu Lee, Ranjay Krishna, et al. 2024. Found in the middle: Calibrating positional attention bias improves long context utilization. arXiv preprint arXiv:2406.16008. find
  11. Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, et al. 2023. A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. arXiv preprint arXiv:2311.05232.
  12. Houcheng Jiang, Junfeng Fang, Ningyu Zhang, Guojun Ma, Mingyang Wan, Xiang Wang, Xiangnan He, and Tat-seng Chua. 2025. Anyedit: Edit any knowledge encoded in language models. ICML. find
  13. Jungo Kasai, Keisuke Sakaguchi, Ronan Le Bras, Akari Asai, Xinyan Yu, Dragomir Radev, Noah A Smith, Yejin Choi, Kentaro Inui, et al. 2024. Realtime qa: what’s the answer right now? Advances in Neural Information Processing Systems, 36.
  14. 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.
  15. 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. In Advances in Neural Information Proc…
  16. Daliang Li, Ankit Singh Rawat, Manzil Zaheer, Xin Wang, Michal Lukasik, Andreas Veit, Felix Yu, and Sanjiv Kumar. 2023. Large language models with controllable working memory. In Findings of the Association for Computational Linguistics: ACL 2023, pages 1774–1793, Toronto, Canada. Association for Co…
  17. Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, et al. 2024a. Deepseek-v3 technical report. arXiv preprint arXiv:2412.19437. find
  18. Bingshuai Liu, Chenyang Lyu, Zijun Min, Zhanyu Wang, Jinsong Su, and Longyue Wang. 2024b. Retrieval-augmented multi-modal chain-of-thoughts reasoning for large language models. Preprint, arXiv:2312.01714.
  19. Nelson F. Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang. 2024c. Lost in the middle: How language models use long contexts. Transactions of the Association for Computational Linguistics, 12:157–173.
  20. Zihan Liu, Wei Ping, Rajarshi Roy, Peng Xu, Chankyu Lee, Mohammad Shoeybi, and Bryan Catanzaro. 2024d. Chatqa: Surpassing gpt-4 on conversational qa and rag. In The Thirty-eighth Annual Conference on Neural Information Processing Systems.
  21. Shayne Longpre, Kartik Perisetla, Anthony Chen, Nikhil Ramesh, Chris DuBois, and Sameer Singh. 2021. Entity-based knowledge conflicts in question answering. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 7052–7063, Online and Punta Cana, Dominican Re…
  22. Yifei Ming, Senthil Purushwalkam, Shrey Pandit, Zixuan Ke, Xuan-Phi Nguyen, Caiming Xiong, and Shafiq Joty. 2025. Faitheval: Can your language model stay faithful to context, even if ”the moon is made of marshmallows”. In The Thirteenth International Conference on Learning Representations. find
  23. Xuan-Phi Nguyen, Shrey Pandit, Senthil Purushwalkam, Austin Xu, Hailin Chen, Yifei Ming, Zixuan Ke, Silvio Savarese, Caiming Xong, and Shafiq Joty. 2024. Sfr-rag: Towards contextually faithful llms. arXiv preprint arXiv:2409.09916. — [arXiv:2409.09916](https://arxiv.org/abs/2409.09916) find
  24. OpenAI. 2023. Gpt-4 technical report. arXiv:2303.08774.
  25. Preprint, Pranav Rajpurkar, Jian Zhang, Konstantin Lopyrev, and Percy Liang. 2016. Squad: 100,000+ questions for machine comprehension of text. Preprint, arXiv:1606.05250.
  26. Weijia Shi, Xiaochuang Han, Mike Lewis, Yulia Tsvetkov, Luke Zettlemoyer, and Scott Wen-tau Yih. 2023. Trusting your evidence: Hallucinate less with context-aware decoding. arXiv preprint arXiv:2305.14739.
  27. Hexiang Tan, Fei Sun, Wanli Yang, Yuanzhuo Wang, Qi Cao, and Xueqi Cheng. 2024. Blinded by generated contexts: How language models merge generated and retrieved contexts when knowledge conflicts? In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Lo…
  28. Harsh Trivedi, Niranjan Balasubramanian, Tushar Khot, and Ashish Sabharwal. 2022. Musique: Multi-hop questions via single-hop question composition. Preprint, arXiv:2108.00573. — [arXiv:2108.00573](https://arxiv.org/abs/2108.00573) find
  29. Song Wang, Yaochen Zhu, Haochen Liu, Zaiyi Zheng, Chen Chen, and Jundong Li. 2024. Knowledge editing for large language models: A survey. ACM Computing Surveys, 57(3):1–37.
  30. Kevin Wu, Eric Wu, and James Zou. 2024. How faithful are rag models? quantifying the tug-of-war between rag and llms’ internal prior. arXiv e-prints, pages arXiv–2404. find
  31. Jian Xie, Kai Zhang, Jiangjie Chen, Renze Lou, and Yu Su. 2024. Adaptive chameleon or stubborn sloth: Revealing the behavior of large language models in knowledge conflicts. In The Twelfth International Conference on Learning Representations.
  32. Peng Xu, Wei Ping, Xianchao Wu, Chejian Xu, Zihan Liu, Mohammad Shoeybi, and Bryan Catanzaro. 2024a. Chatqa 2: Bridging the gap to proprietary llms in long context and rag capabilities. arXiv preprint arXiv:2407.14482. — [arXiv:2407.14482](https://arxiv.org/abs/2407.14482) find
  33. Rongwu Xu, Zehan Qi, Zhijiang Guo, Cunxiang Wang, Hongru Wang, Yue Zhang, and Wei Xu. 2024b. Knowledge conflicts for LLMs: A survey. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 8541– 8565, Miami, Florida, USA. Association for Computational Linguis…
  34. Jiahao Ying, Yixin Cao, Kai Xiong, Long Cui, Yidong He, and Yongbin Liu. 2024. Intuitive or dependent? investigating LLMs’ behavior style to conflicting prompts. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4221–4246, Bangk… find
  35. Wenhao Yu, Dan Iter, Shuohang Wang, Yichong Xu, Mingxuan Ju, Soumya Sanyal, Chenguang Zhu, Michael Zeng, and Meng Jiang. 2023. Generate rather than retrieve: Large language models are strong context generators. In The Eleventh International Conference on Learning Representations.
  36. Xiaowei Yuan, Zhao Yang, Yequan Wang, Shengping Liu, Jun Zhao, and Kang Liu. 2024. Discerning and resolving knowledge conflicts through adaptive decoding with contextual information-entropy constraint. In Findings of the Association for Computational Linguistics: ACL 2024, pages 3903–3922, Bangkok,… find
  37. Zheng Yuan, Hao Chen, Zijin Hong, Qinggang Zhang, Feiran Huang, and Xiao Huang. 2025. Knapsack optimization-based schema linking for llm-based text-to-sql generation. arXiv preprint arXiv:2502.12911. — [arXiv:2502.12911](https://arxiv.org/abs/2502.12911) find
  38. Qinggang Zhang, Shengyuan Chen, Yuanchen Bei, Zheng Yuan, Huachi Zhou, Zijin Hong, Junnan Dong, Hao Chen, Yi Chang, and Xiao Huang. 2025. A survey of graph retrieval-augmented generation for customized large language models. arXiv preprint arXiv:2501.13958.
  39. Qinggang Zhang, Junnan Dong, Hao Chen, Daochen Zha, Zailiang Yu, and Xiao Huang. 2024. Knowgpt: Knowledge graph based prompting for large language models. Advances in Neural Information Processing Systems, 37:6052–6080. find
  40. Shengyu Zhang, Linfeng Dong, Xiaoya Li, Sen Zhang, Xiaofei Sun, Shuhe Wang, Jiwei Li, Runyi Hu, Tianwei Zhang, Fei Wu, et al. 2023. Instruction tuning for large language models: A survey. arXiv preprint arXiv:2308.10792.
  41. Chuang Zhou, Jiahe Du, Huachi Zhou, Hao Chen, Feiran Huang, and Xiao Huang. 2025. Text-attributed graph learning with coupled augmentations. In Proceedings of the 31st International Conference on Computational Linguistics, pages 10865–10876.
  42. Huachi Zhou, Shuang Zhou, Hao Chen, Ninghao Liu, Fan Yang, and Xiao Huang. 2024. Enhancing explainable rating prediction through annotated macro concepts. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 11736–11748. find
  43. Wenxuan Zhou, Sheng Zhang, Hoifung Poon, and Muhao Chen. 2023. Context-faithful prompting for large language models. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 14544–14556, Singapore. Association for Computational Linguistics.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zhang-2025-faithfulrag-fact-level-conflict-modeling/