Elamur Research⌕
☰
home / papers / 2026 / Knowledge-to-Verification: Exploring RLVR for LLMs in Knowledge-Intensive Domains

Knowledge-to-Verification: Exploring RLVR for LLMs in Knowledge-Intensive Domains

Paper · 2026 · Zhonghang Yuan, Zhefan Wang, Fang Hu et al.

Summary

RLVR has improved LLM reasoning mainly in math and code, where answers can be checked automatically. Fields such as agriculture, law and medicine lack such checkable data, and answer-only rewards are sparse and tolerate flawed reasoning. The authors propose K2V (Knowledge-to-Verification) to address both gaps. K2V uses GraphGen to build a knowledge graph from raw domain corpora through LLM-driven NER and relation extraction. It then samples entity–relation chains of five elements (quintuples), hides one entity, and has an LLM rewrite the chain as a fill-in-the-blank question. The hidden entity serves as the ground truth, which a rule-based checker can verify. To assess reasoning, each question gets a checklist of yes/no criteria derived from general templates. A judge LLM scores the model's reasoning against the checklist, and the pass rate becomes a dense reward. That reward counts only when the final answer is correct, which limits reward hacking. Experiments with Qwen2.5 and Llama3 backbones trained with DAPO show gains on domain benchmarks while general skills stay largely intact. The work offers a practical way to bring RLVR to specialist fields.

Key points

  • Evaluation benchmarks:
    • Domain: SeedBench (agriculture), LawBench and MedQA.
    • General: BBH, MATH-500, GSM8K, AIME2024 and GPQA-Diamond.
  • Models in the pipeline:
    • Qwen2.5-72B-Instruct is the data generator; Qwen2.5-7B-Instruct is the judge.
    • The policy models are small to mid-sized, drawn from the Qwen and Llama families at 3B and 8B parameters.
  • Generator size matters little: With smaller generators (14B or 7B), Qwen2.5-3B scores fall only slightly. With the 72B generator, the model reaches 62.82 on SeedBench, 43.27 on LawBench and 78.45 on MedQA.
  • Ablations:
    • Plain SFT on the synthesized QA pairs, without RL, causes a sharp performance drop.
    • Results are stable across answer-reward values α ∈ {2, 4, 6, 8, 10}.
  • Synthesis cost: About 281.59 H100 GPU hours in total with the 72B generator, of which 201.25 hours go to building the knowledge graph.

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

🔬Shanghai Artificial Intelligence LaboratoryShanghai · China152 works · 13,164 citations🎓Shanghai Innovation InstituteShanghai · China34 works · 325 citations

 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 worksComputer Vision and Pattern Recognition · 3,486 works

🧭 Topics 2Research subjects and languages the work is about.Language models & pretraining · 3,894 worksReasoning · 1,685 works

Cite

Cite as Zhonghang Yuan, Zhefan Wang, Fang Hu, Zihong Chen, Jinzhe Li, Gang Li, Jie Ying, Huanjun Kong, Songyang Zhang and Nanqing Dong. 2026. Knowledge-to-Verification: Exploring RLVR for LLMs in Knowledge-Intensive Domains. Proceedings of the 64th 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/v1/2026.acl-long.1891

@inproceedings{yuan2026knowledge,
  title = {{Knowledge-to-Verification: Exploring RLVR for LLMs in Knowledge-Intensive Domains}},
  author = {Zhonghang Yuan, Zhefan Wang, Fang Hu, Zihong Chen, Jinzhe Li, Gang Li, Jie Ying, Huanjun Kong, Songyang Zhang and Nanqing Dong},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1891},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1891}
}

References

60 references, 28 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Sultan Alneyadi, Elankayer Sithirasenan, and Vallipu-ram Muthukkumarasamy. 2016. A survey on data leakage prevention systems. Journal of Network and Computer Applications, 62:137–152. find
  2. Yanda Chen, Joe Benton, Ansh Radhakrishnan, Jonathan Uesato, Carson Denison, John Schulman, Arushi Somani, Peter Hase, Misha Wagner, Fabien Roger, Vladimir Mikulik, Samuel R. Bowman, Jan Leike, Jared Kaplan, and Ethan Perez. 2025a. Reasoning models don’t always say what they think. CoRR, abs/2505.05…
  3. Zihong Chen, Wanli Jiang, Jinzhe Li, Zhonghang Yuan, Huanjun Kong, Wanli Ouyang, and Nanqing Dong. 2025b. Graphgen: Enhancing supervised fine-tuning for llms with knowledge-driven synthetic data generation. arXiv preprint arXiv:2505.20416. — [arXiv:2505.20416](https://arxiv.org/abs/2505.20416) find
  4. 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. CoRR, abs/2110.14168.
  5. OpenCompass Contributors. 2023. Opencompass: A universal evaluation platform for foundation models. https://github.com/open-compass/ opencompass. ↗
  6. Thomas H Cormen, Charles E Leiserson, Ronald L Rivest, and Clifford Stein. 2022. Introduction to algorithms. MIT press. find
  7. Ganqu Cui, Lifan Yuan, Zefan Wang, Hanbin Wang, Wendi Li, Bingxiang He, Yuchen Fan, Tianyu Yu, Qixin Xu, Weize Chen, Jiarui Yuan, Huayu Chen, Kaiyan Zhang, Xingtai Lv, Shuo Wang, Yuan Yao, Xu Han, Hao Peng, Yu Cheng, and 4 others. 2025. Process reinforcement through implicit rewards. CoRR, abs/2502.…
  8. Bhishma Dedhia, Yuval Kansal, and Niraj K Jha. 2025. Bottom-up domain-specific superintelligence: A reliable knowledge graph is what we need. arXiv preprint arXiv:2507.13966. — [arXiv:2507.13966](https://arxiv.org/abs/2507.13966) find
  9. Yann Dubois, Chen Xuechen Li, Rohan Taori, Tianyi Zhang, Ishaan Gulrajani, Jimmy Ba, Carlos Guestrin, Percy S Liang, and Tatsunori B Hashimoto. 2023. Alpacafarm: A simulation framework for methods that learn from human feedback. Advances in Neural Information Processing Systems, 36:30039–30069.
  10. Zhiwei Fei, Xiaoyu Shen, Dawei Zhu, Fengzhe Zhou, Zhuo Han, Alan Huang, Songyang Zhang, Kai Chen, Zhixin Yin, Zongwen Shen, Jidong Ge, and Vincent Ng. 2024. LawBench: Benchmarking legal knowledge of large language models. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language…
  11. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, Amy Yang, Angela Fan, Anirudh Goyal, Anthony Hartshorn, Aobo Yang, Archi Mitra, Archie Sravankumar, Artem Korenev, Arthur Hinsvark, and 542 o… — The Llama 3 Herd of Models
  12. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, Ruoyu Zhang, Shirong Ma, Xiao Bi, Xiaokang Zhang, Xingkai Yu, Yu Wu, Z. F. Wu, Zhibin Gou, Zhihong Shao, Zhuoshu Li, Ziyi Gao, Aixin Liu, and 175 others. 2025a. Deepseek-r1: incentivizes reasoning in llms through re… find
  13. Yiduo Guo, Zhen Guo, Chuanwei Huang, Zi-Ang Wang, Zekai Zhang, Haofei Yu, Huishuai Zhang, and Yikang Shen. 2025b. Synthetic data rl: Task definition is all you need. arXiv preprint arXiv:2505.17063. — [arXiv:2505.17063](https://arxiv.org/abs/2505.17063) find
  14. Jujie He, Jiacai Liu, Chris Yuhao Liu, Rui Yan, Chaojie Wang, Peng Cheng, Xiaoyu Zhang, Fuxiang Zhang, Jiacheng Xu, Wei Shen, Siyuan Li, Liang Zeng, Tianwen Wei, Cheng Cheng, Bo An, Yang Liu, and Yahui Zhou. 2025. Skywork open reasoner 1 technical report. CoRR, abs/2505.22312. find
  15. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2020. Measuring massive multitask language understanding. arXiv preprint arXiv:2009.03300.
  16. Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. 2021. Measuring mathematical problem solving with the math dataset. arXiv preprint arXiv:2103.03874.
  17. Aidan Hogan, Eva Blomqvist, Michael Cochez, Claudia d’Amato, Gerard de Melo, Claudio Gutierrez, Sabrina Kirrane, José Emilio Labra Gayo, Roberto Navigli, Sebastian Neumaier, Axel-Cyrille Ngonga Ngomo, Axel Polleres, Sabbir M. Rashid, Anisa Rula, Lukas Schmelzeisen, Juan F. Sequeda, Steffen Staab, an… — Knowledge Graphs
  18. Matthew Honnibal, Ines Montani, Sofie Van Lan-deghem, and Adriane Boyd. 2020. spaCy: Industrial-strength Natural Language Processing in Python. find
  19. Jingcheng Hu, Yinmin Zhang, Qi Han, Daxin Jiang, Xiangyu Zhang, and Heung-Yeung Shum. 2025. Open-reasoner-zero: An open source approach to scaling up reinforcement learning on the base model. arXiv preprint arXiv:2503.24290.
  20. Hugging Face. 2025. Open r1: A fully open reproduction of deepseek-r1. find
  21. Shaoxiong Ji, Shirui Pan, Erik Cambria, Pekka Marttinen, and Philip S Yu. 2021. A survey on knowledge graphs: Representation, acquisition, and applications. find
  22. IEEE transactions on neural networks and learning systems, 33(2):494–514. find
  23. Di Jin, Eileen Pan, Nassim Oufattole, Wei-Hung Weng, Hanyi Fang, and Peter Szolovits. 2021. What disease does this patient have? a large-scale open domain question answering dataset from medical exams. Applied Sciences, 11(14):6421.
  24. Imed Keraghel, Stanislas Morbieu, and Mohamed Nadif. 2024. A survey on recent advances in named entity recognition. arXiv preprint arXiv:2401.10825. — [arXiv:2401.10825](https://arxiv.org/abs/2401.10825) find
  25. Seongyun Lee, Hyunjae Kim, and Jaewoo Kang. 2023. Liquid: A framework for list question answering dataset generation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 13014–13024. find
  26. Haonan Li, Yixuan Zhang, Fajri Koto, Yifei Yang, Hai Zhao, Yeyun Gong, Nan Duan, and Timothy Baldwin. 2024. Cmmlu: Measuring massive multitask language understanding in chinese. In Findings of the Association for Computational Linguistics: ACL 2024, pages 11260–11285.
  27. Hunter Lightman, Vineet Kosaraju, Yuri Burda, Harrison Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, and Karl Cobbe. 2023. Let’s verify step by step. In The Twelfth International Conference on Learning Representations.
  28. Zichen Liu, Changyu Chen, Wenjun Li, Penghui Qi, Tianyu Pang, Chao Du, Wee Sun Lee, and Min Lin. 2025. Understanding r1-zero-like training: A critical perspective. arXiv preprint arXiv:2503.20783.
  29. Yuxing Lu, Wei Wu, Xukai Zhao, Rui Peng, and Jinzhuo Wang. Karma: Leveraging multi-agent llms for automated knowledge graph enrichment. In The Thirty-ninth Annual Conference on Neural Information Processing Systems. find
  30. Michael Luo, Sijun Tan, Roy Huang, Ameen Patel, Alpay Ariyak, Qingyang Wu, Xiaoxiang Shi, Rachel Xin, Colin Cai, and Maurice Weber. 2025a. Deepcoder: A fully open-source 14b coder at o3-mini level. Notion Blog. find
  31. Michael Luo, Sijun Tan, Justin Wong, Xiaoxiang Shi, William Y. Tang, Manan Roongta, Colin Cai, Jeffrey Luo, Li Erran Li, Raluca Ada Popa, and Ion Stoica. 2025b. Deepscaler: Surpassing o1-preview with a 1.5b model by scaling rl. Notion Blog. find
  32. Xueguang Ma, Qian Liu, Dongfu Jiang, Ge Zhang, Zejun Ma, and Wenhu Chen. 2025. General-reasoner: Advancing llm reasoning across all domains. arXiv preprint arXiv:2505.14652.
  33. Carney, Alex Iftimie, Alex Karpenko, Alex Tachard Passos, Alexander Neitz, Alexander Prokofiev, Alexander Wei, Allison Tam, and 244 others. 2024. Openai o1 system card. Preprint, arXiv:2412.16720.
  34. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul F. Christiano, Jan Leike, and Ryan Lowe. 2022.… — Training language models to follow instructions with human feedback
  35. Shirui Pan, Linhao Luo, Yufei Wang, Chen Chen, Jiapu Wang, and Xindong Wu. 2024. Unifying large language models and knowledge graphs: A roadmap. IEEE Transactions on Knowledge and Data Engineering, 36(7):3580–3599.
  36. Miao Peng, Nuo Chen, Zongrui Suo, and Jia Li. 2025. Rewarding graph reasoning process makes llms more generalized reasoners. In Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 2, pages 2257–2268. find
  37. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R Bowman. 2024. Gpqa: A graduate-level google-proof q&a benchmark. In First Conference on Language Modeling.
  38. Maohao Shen, Guangtao Zeng, Zhenting Qi, Zhang-Wei Hong, Zhenfang Chen, Wei Lu, Gregory Wornell, Subhro Das, David Cox, and Chuang Gan. 2025. Satori: Reinforcement learning with chain-of-action-thought enhances llm reasoning via autoregressive search. In International Conference on Machine Learning. find
  39. Guangming Sheng, Chi Zhang, Zilingfeng Ye, Xibin Wu, Wang Zhang, Ru Zhang, Yanghua Peng, Haibin Lin, and Chuan Wu. 2025. Hybridflow: A flexible and efficient rlhf framework. In Proceedings of the Twentieth European Conference on Computer Systems, pages 1279–1297.
  40. Zafir Stojanovski, Oliver Stanley, Joe Sharratt, Richard Jones, Abdulhakeem Adefioye, Jean Kaddour, and Andreas Köpf. 2025. Reasoning gym: Reasoning environments for reinforcement learning with verifiable rewards. arXiv preprint arXiv:2505.24760. — [arXiv:2505.24760](https://arxiv.org/abs/2505.24760) find
  41. MAA. 2024. American invitational mathematics examination - aime. Accessed on: 2025-09-21. find
  42. Zhenpeng Su, Leiyu Pan, Xue Bai, Dening Liu, Guanting Dong, Jiaming Huang, Wenping Hu, and Guorui Zhou. 2025. Klear-reasoner: Advancing reasoning capability via gradient-preserving clipping policy optimization. arXiv preprint arXiv:2508.07629. — [arXiv:2508.07629](https://arxiv.org/abs/2508.07629) find
  43. OpenAI, :, Aaron Jaech, Adam Kalai, Adam Lerer, Adam Richardson, Ahmed El-Kishky, Aiden Low, Alec Helyar, Aleksander Madry, Alex Beutel, Alex Mujeen Sung, Minbyul Jeong, Yonghwa Choi, Donghyeon Kim, Jinhyuk Lee, and Jaewoo Kang. 2022. Bern2: an advanced neural biomedical named entity recognition and… find
  44. Mirac Suzgun, Nathan Scales, Nathanael Schärli, Sebastian Gehrmann, Yi Tay, Hyung Won Chung, Aakanksha Chowdhery, Quoc Le, Ed Chi, Denny Zhou, and Jason Wei. 2023. Challenging BIG-bench tasks and whether chain-of-thought can solve them. In Findings of the Association for Computational Linguistics: A…
  45. Liang Yao, Jiazhen Peng, Chengsheng Mao, and Yuan Luo. 2025. Exploring large language models for knowledge graph completion. In ICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 1–5. IEEE. find
  46. Yizhong Wang, Yeganeh Kordi, Swaroop Mishra, Alisa Liu, Noah A. Smith, Daniel Khashabi, and Hannaneh Hajishirzi. 2023. Self-instruct: Aligning language models with self-generated instructions. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long… find
  47. Zengzhi Wang, Fan Zhou, Xuefeng Li, and Pengfei Liu. 2025. Octothinker: Mid-training incentivizes reinforcement learning scaling. arXiv preprint arXiv:2506.20512. — [arXiv:2506.20512](https://arxiv.org/abs/2506.20512) find
  48. Liang Wen, Yunke Cai, Fenrui Xiao, Xin He, Qi An, Zhenyu Duan, Yimin Du, Junchen Liu, Lifu Tang, Xiaowei Lv, Haosheng Zou, Yongchao Deng, Shousheng Jia, and Xiangzheng Zhang. 2025. Light-r1: Curriculum SFT, DPO and RL for long COT from scratch and beyond. In Proceedings of the 63rd Annual Meeting of…
  49. Can Xu, Qingfeng Sun, Kai Zheng, Xiubo Geng, Pu Zhao, Jiazhan Feng, Chongyang Tao, Qingwei Lin, and Daxin Jiang. 2024. Wizardlm: Empowering large pre-trained language models to follow complex instructions. In The Twelfth International Conference on Learning Representations.
  50. 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. 2025a. Qwen3 technical report. Preprint, arXiv:2505.09388.
  51. An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, and 22 others. 2024. Qwen2.5 technical report. CoRR, abs/2412.1511… find
  52. Fan Yang, Huanjun Kong, Jie Ying, Zihong Chen, Tao Luo, Wanli Jiang, Zhonghang Yuan, Zhefan Wang, Zhaona Ma, Shikuan Wang, Wanfeng Ma, Xiaoyi Wang, Xiaoying Li, Zhengyin Hu, Xiaodong Ma, Minguo Liu, Xiqing Wang, Fan Chen, and Nanqing Dong. 2025b. Seedllm·rice: A large language Asaf Yehudai, Boaz Car… — [arXiv:2401.14367](https://arxiv.org/abs/2401.14367) find
  53. Jie Ying, Zihong Chen, Zhefan Wang, Wanli Jiang, Chenyang Wang, Zhonghang Yuan, Haoyang Su, Huanjun Kong, Fan Yang, and Nanqing Dong. 2025. SeedBench: A multi-task benchmark for evaluating large language models in seed science. In Proceedings of the 63rd Annual Meeting of the Association for Computa… find
  54. Ping Yu, Jack Lanchantin, Tianlu Wang, Weizhe Yuan, Olga Golovneva, Ilia Kulikov, Sainbayar Sukhbaatar, Jason Weston, and Jing Xu. 2025a. Cot-self-instruct: Building high-quality synthetic prompts for reasoning and non-reasoning tasks. arXiv preprint arXiv:2507.23751. — [arXiv:2507.23751](https://arxiv.org/abs/2507.23751) find
  55. Qiying Yu, Zheng Zhang, Ruofei Zhu, Yufeng Yuan, Xiaochen Zuo, Yu Yue, Weinan Dai, Tiantian Fan, Gaohong Liu, Lingjun Liu, Xin Liu, Haibin Lin, Zhiqi Lin, Bole Ma, Guangming Sheng, Yuxuan Tong, Chi Zhang, Mofan Zhang, Wang Zhang, and 16 others. 2025b. Dapo: An open-source llm reinforcement learning…
  56. Shengbin Yue, Wei Chen, Siyuan Wang, Bingxuan Li, Chenchen Shen, Shujun Liu, Yuxuan Zhou, Yao Xiao, Song Yun, Xuanjing Huang, and Zhongyu Wei. 2023. Disc-lawllm: Fine-tuning large language models for intelligent legal services. Preprint, arXiv:2309.11325.
  57. Weihao Zeng, Yuzhen Huang, Qian Liu, Wei Liu, Keqing He, Zejun Ma, and Junxian He. 2025. Simplerl-zoo: Investigating and taming zero reinforcement learning for open base models in the wild. arXiv preprint arXiv:2503.18892.
  58. Xiaoyan Zhao, Yang Deng, Min Yang, Lingzhi Wang, Rui Zhang, Hong Cheng, Wai Lam, Ying Shen, and Ruifeng Xu. 2024. A comprehensive survey on relation extraction: Recent advances and new frontiers. ACM Computing Surveys, 56(11):1–39. find
  59. Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric P. Xing, Hao Zhang, Joseph E. Gonzalez, and Ion Stoica. 2023. Judging llm-as-a-judge with mt-bench and chatbot arena. In Advances in Neural Information Processing Systems 36: A…
  60. Xiangxin Zhou, Zichen Liu, Anya Sims, Haonan Wang, Tianyu Pang, Chongxuan Li, Liang Wang, Min Lin, and Chao Du. 2025. Reinforcing general reasoning without verifiers. arXiv preprint arXiv:2505.21493. — [arXiv:2505.21493](https://arxiv.org/abs/2505.21493) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/yuan-2026-knowledge-verification-exploring-rlvr-llms/