---
title: "Rethinking Sample Polarity in Reinforcement Learning with"
description: "From these findings they build A3PO. 42 references, cited by 4 papers here. Among them: 0."
authors: ["Xinyu Tang", "Yuliang Zhan", "Zhixun Li", "Xin Zhao", "Zhenduo Zhang", "Zujie Wen", "Zhiqiang Zhang", "Jun Zhou"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2026.acl-long.134/"
doi: "10.18653/v1/2026.acl-long.134"
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/tang-2026-rethinking-sample-polarity-reinforcement-learning/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/tang-2026-rethinking-sample-polarity-reinforcement-learning/)

[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/4/) / Rethinking Sample Polarity in Reinforcement Learning with Verifiable Rewards

# Rethinking Sample Polarity in Reinforcement Learning with Verifiable Rewards

**Authors:** [Xinyu Tang](https://md.research.elamur.ai/free/en/authors/xinyu-tang/), Yuliang Zhan, [Zhixun Li](https://md.research.elamur.ai/free/en/authors/zhixun-li/), [Xin Zhao](https://md.research.elamur.ai/free/en/authors/xin-zhao/), Zhenduo Zhang, Zujie Wen, [Zhiqiang Zhang](https://md.research.elamur.ai/free/en/authors/zhiqiang-zhang/) and [Jun Zhou](https://md.research.elamur.ai/free/en/authors/jun-zhou/) · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/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.acl-long.134/) · [PDF](https://aclanthology.org/2026.acl-long.134.pdf) · [DOI 10.18653/v1/2026.acl-long.134](https://doi.org/10.18653/v1/2026.acl-long.134) · [OpenAlex](https://openalex.org/W7166832685)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · [4](https://md.research.elamur.ai/free/en/papers/tang-2026-rethinking-sample-polarity-reinforcement-learning/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
This paper asks what positive (correct) and negative (incorrect) rollouts each contribute when large reasoning models are trained with RLVR. The authors train three base models, Qwen2.5-7B-Math, Qwen3-8B-Base and DeepSeek-R1-Distill-Qwen-7B, three ways: on positive samples only (PSR), on negative samples only (NSR), and on both (DAPO). They then compare the training dynamics of each setup.

Learning from correct answers makes the model more confident in reasoning paths it already uses. Entropy falls and responses get shorter. Learning from errors keeps entropy high, produces longer outputs and leads to more novel solution paths. Neither polarity works well alone, so both are needed.

The authors also test reweighting advantages for a whole polarity and for individual tokens, based on token entropy and probability. From these findings they build A3PO. The method adaptively boosts the advantage of low-probability tokens in correct rollouts and high-probability tokens in incorrect ones. A3PO beats GRPO, DAPO and other recent variants on five math, science and coding benchmarks. This offers a principled way to balance exploration and exploitation in RL for reasoning models.

## Key points
- Evaluation covers AIME24, AIME25, MATH500, GPQA and LiveCodeBench. Baselines are GRPO, DAPO, DAPO w/ Fork Tokens, W-REINFORCE and Lp-Reg.
- Average scores for A3PO versus DAPO:
  - Qwen3-8B-Base: 48.7 vs 44.1
  - DeepSeek-R1-Distill-Qwen-7B: 63.4 vs 60.1
  - Qwen2.5-7B-Math: 40.4 vs 36.3

  All gains over baselines are statistically significant (t-test, p \< 0.05).
- On AIME problems at step 100, the share of new reasoning paths was 38.3% for NSR, 20.0% for PSR and 10.0% for DAPO.
- With Qwen3-8B-Base, PSR-only training led to reward hacking: the model guessed answers instead of reasoning step by step. The earlier claim that negative-only training widens the reasoning boundary held only for Qwen2.5-7B-Math.
- Ablation on Qwen3-8B-Base (math average 53.2 for the full method):
  - Removing both token-shaping components drops it to 48.3.
  - Removing the adaptive schedule drops it to 51.8.

  Code is available at github.com/RUCAIBox/Sample-Polarity-of-RLVR.

*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

- [Renmin University of China](https://md.research.elamur.ai/free/en/universities/renmin-university-of-china/) — University, Beijing, China · 114 works · 16,118 citations
- [Chinese University of Hong Kong](https://md.research.elamur.ai/free/en/universities/chinese-university-of-hong-kong/) — University, Hong Kong, China · 126 works · 14,073 citations
- [Ant Group](https://md.research.elamur.ai/free/en/companies/ant-group/) — Company, Hangzhou, China · 42 works · 1,131 citations

📚 Shelves (4): [🖥️ 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

🧭 Topics (1): [🧮 Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/) · 1,426 works

## Cite

Cite as: Xinyu Tang, Yuliang Zhan, Zhixun Li, Xin Zhao, Zhenduo Zhang, Zujie Wen, Zhiqiang Zhang and Jun Zhou. 2026. Rethinking Sample Polarity in Reinforcement Learning with Verifiable Rewards. 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.

```bibtex
@inproceedings{tang2026rethinking,
  title = {{Rethinking Sample Polarity in Reinforcement Learning with Verifiable Rewards}},
  author = {Xinyu Tang, Yuliang Zhan, Zhixun Li, Xin Zhao, Zhenduo Zhang, Zujie Wen, Zhiqiang Zhang and Jun Zhou},
  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.134},
  url = {https://doi.org/10.18653/v1/2026.acl-long.134}
}
```

- [← previous Query-Efficient Agentic Graph Extraction Attacks on GraphRAG Systems](https://md.research.elamur.ai/free/en/papers/yang-2026-query-efficient-agentic-graph-extraction/)
- [next → SAHM: A Benchmark for Arabic Financial and Shari'ah-Compliant Reasoning](https://md.research.elamur.ai/free/en/papers/elbadry-2026-sahm-benchmark-arabic-financial-shari/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/4/)

## References

42 references, 17 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Aili Chen, Aonian Li, Bangwei Gong, Binyang Jiang, Bo Fei, Bo Yang, Boji Shan, Changqing Yu, Chao Wang, Cheng Zhu, Chengjun Xiao, Chengyu Du, Chi Zhang, Chu Qiao, Chunhao Zhang, Chunhui Du, Congchao Guo, Da Chen, Deming Ding, Dianjun Sun, Dong Li, Enwei Jiao, Haigang Zhou, Haimo Zhang, Han Ding, Hao… — [MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention](https://md.research.elamur.ai/free/en/papers/minimax-2025-minimax-m1-scaling-test/)
2. Zhipeng Chen, Xiaobo Qin, Youbin Wu, Yue Ling, Qinghao Ye, Wayne Xin Zhao, and Guang Shi. 2025b. Pass@k training for adaptively balancing exploration and exploitation of large reasoning models. CoRR, abs/2508.10751. — [arXiv:2508.10751](https://arxiv.org/abs/2508.10751) [find](https://scholar.google.com/scholar?q=Zhipeng+Chen%2C+Xiaobo+Qin%2C+Youbin+Wu%2C+Yue+Ling%2C+Qinghao+Ye%2C+Wayne+Xin+Zhao%2C+and+Guang+Shi%2E+2025b%2E+Pass%40k+training+for+adaptively+balancing+exploration+and+exploitation+of+large+reasoning+models%2E+CoRR%2C+)
3. Zhipeng Chen, Xiaobo Qin, Wayne Xin Zhao, Youbin Wu, and Ji-Rong Wen. 2026. Adaptive ability decomposing for unlocking large reasoning model effective reinforcement learning. CoRR, abs/2602.00759. — [arXiv:2602.00759](https://arxiv.org/abs/2602.00759) [find](https://scholar.google.com/scholar?q=Zhipeng+Chen%2C+Xiaobo+Qin%2C+Wayne+Xin+Zhao%2C+Youbin+Wu%2C+and+Ji%2DRong+Wen%2E+2026%2E+Adaptive+ability+decomposing+for+unlocking+large+reasoning+model+effective+reinforcement+learning%2E+CoRR%2C+abs%2F2602%2E00759%2E+%E2%80%94+%5B)
4. DeepSeek-AI, Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, Xiaokang Zhang, Xingkai Yu, Yu Wu, Z. F. Wu, Zhibin Gou, Zhihong Shao, Zhuoshu Li, Ziyi Gao, Aixin Liu, Bing Xue, Bingxuan Wang, Bochao Wu, Bei Feng, Chengda Lu, Chengg… — [Rethinking Entropy Interventions in RLVR: An Entropy Change Perspective](https://md.research.elamur.ai/free/en/papers/hao-2026-rethinking-entropy-interventions-rlvr-entropy/)
5. 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](https://md.research.elamur.ai/free/en/papers/hendrycks-2021-measuring-mathematical-problem-solving-math/). In NeurIPS Datasets and Benchmarks.
6. Guanhua Huang, Tingqiang Xu, Mingze Wang, Qi Yi, Xue Gong, Siheng Li, Ruibin Xiong, Kejiao Li, Yuhao Jiang, and Bo Zhou. 2025. Low-probability tokens sustain exploration in reinforcement learning with verifiable reward. CoRR, abs/2510.03222. [find](https://scholar.google.com/scholar?q=Guanhua+Huang%2C+Tingqiang+Xu%2C+Mingze+Wang%2C+Qi+Yi%2C+Xue+Gong%2C+Siheng+Li%2C+Ruibin+Xiong%2C+Kejiao+Li%2C+Yuhao+Jiang%2C+and+Bo+Zhou%2E+2025%2E+Low%2Dprobability+tokens+sustain+exploration+in+reinforcement+learning+with)
7. Naman Jain, King Han, Alex Gu, Wen-Ding Li, Fanjia Yan, Tianjun Zhang, Sida Wang, Armando Solar-Lezama, Koushik Sen, and Ion Stoica. 2025. [Livecodebench: Holistic and contamination free evaluation of large language models for code](https://md.research.elamur.ai/free/en/papers/jain-2024-livecodebench-holistic-contamination-free-evaluation/). In ICLR. OpenReview.net.
8. 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](https://md.research.elamur.ai/free/en/papers/kwon-2023-efficient-memory-management-large-language/). In SOSP, pages 611–626. ACM.
9. Yifan Li, Zhenghao Chen, Ziheng Wu, Kun Zhou, Ruipu Luo, Can Zhang, Zhentao He, Yufei Zhan, Wayne Xin Zhao, and Minghui Qiu. 2025a. Unleashing perception-time scaling to multimodal reasoning models. CoRR, abs/2510.08964. — [arXiv:2510.08964](https://arxiv.org/abs/2510.08964) [find](https://scholar.google.com/scholar?q=Yifan+Li%2C+Zhenghao+Chen%2C+Ziheng+Wu%2C+Kun+Zhou%2C+Ruipu+Luo%2C+Can+Zhang%2C+Zhentao+He%2C+Yufei+Zhan%2C+Wayne+Xin+Zhao%2C+and+Minghui+Qiu%2E+2025a%2E+Unleashing+perception%2Dtime+scaling+to+multimodal+reasoning+models%2E+C)
10. Yifan Li, Yukai Gu, Yingqian Min, Zikang Liu, Yifan Du, Kun Zhou, Min Yang, Wayne Xin Zhao, and Minghui Qiu. 2025b. Beyond the last frame: Process-aware evaluation for generative video reasoning. arXiv preprint arXiv:2512.24952. [find](https://scholar.google.com/scholar?q=Yifan+Li%2C+Yukai+Gu%2C+Yingqian+Min%2C+Zikang+Liu%2C+Yifan+Du%2C+Kun+Zhou%2C+Min+Yang%2C+Wayne+Xin+Zhao%2C+and+Minghui+Qiu%2E+2025b%2E+Beyond+the+last+frame%3A+Process%2Daware+evaluation+for+generative+video+reasoning%2E+arXi)
11. Yifan Li, Kun Zhou, Xin Zhao, Lei Fang, and Jirong Wen. 2026. Analyzing and mitigating object hallucination: A training bias perspective. In AAAI, pages 6636–6643. AAAI Press. [find](https://scholar.google.com/scholar?q=Yifan+Li%2C+Kun+Zhou%2C+Xin+Zhao%2C+Lei+Fang%2C+and+Jirong+Wen%2E+2026%2E+Analyzing+and+mitigating+object+hallucination%3A+A+training+bias+perspective%2E+In+AAAI%2C+pages+6636%E2%80%936643%2E+AAAI+Press%2E)
12. Yizhi Li, Qingshui Gu, Zhoufutu Wen, Ziniu Li, Tianshun Xing, Shuyue Guo, Tianyu Zheng, Xin Zhou, Xingwei Qu, Wangchunshu Zhou, Zheng Zhang, Wei Shen, Qian Liu, Chenghua Lin, Jian Yang, Ge Zhang, and Wenhao Huang. 2025c. Treepo: Bridging the gap of policy optimization and efficacy and inference effi… [find](https://scholar.google.com/scholar?q=Yizhi+Li%2C+Qingshui+Gu%2C+Zhoufutu+Wen%2C+Ziniu+Li%2C+Tianshun+Xing%2C+Shuyue+Guo%2C+Tianyu+Zheng%2C+Xin+Zhou%2C+Xingwei+Qu%2C+Wangchunshu+Zhou%2C+Zheng+Zhang%2C+Wei+Shen%2C+Qian+Liu%2C+Chenghua+Lin%2C+Jian+Yang%2C+Ge+Zhang%2C+and+)
13. Mingjie Liu, Shizhe Diao, Ximing Lu, Jian Hu, Xin Dong, Yejin Choi, Jan Kautz, and Yi Dong. 2025. Prorl: Prolonged reinforcement learning expands reasoning boundaries in large language models. CoRR, abs/2505.24864. [find](https://scholar.google.com/scholar?q=Mingjie+Liu%2C+Shizhe+Diao%2C+Ximing+Lu%2C+Jian+Hu%2C+Xin+Dong%2C+Yejin+Choi%2C+Jan+Kautz%2C+and+Yi+Dong%2E+2025%2E+Prorl%3A+Prolonged+reinforcement+learning+expands+reasoning+boundaries+in+large+language+models%2E+CoRR%2C+a)
14. Yurou Liu, Mingyang Li, Xinyuan Zhu, Rui Jiao, Yiming Dong, Xinyu Tang, Yang Liu, Jieping Ye, Bing Su, and Zheng Wang. 2026. Drugtrail: Explainable drug discovery via structured reasoning and drugga-bility-tailored preference optimization. In The Fourteenth International Conference on Learning Repre… [find](https://scholar.google.com/scholar?q=Yurou+Liu%2C+Mingyang+Li%2C+Xinyuan+Zhu%2C+Rui+Jiao%2C+Yiming+Dong%2C+Xinyu+Tang%2C+Yang+Liu%2C+Jieping+Ye%2C+Bing+Su%2C+and+Zheng+Wang%2E+2026%2E+Drugtrail%3A+Explainable+drug+discovery+via+structured+reasoning+and+drugga%2Db)
15. Ilya Loshchilov and Frank Hutter. 2019. Decoupled weight decay regularization. In ICLR (Poster). Open-Review.net. [find](https://scholar.google.com/scholar?q=Ilya+Loshchilov+and+Frank+Hutter%2E+2019%2E+Decoupled+weight+decay+regularization%2E+In+ICLR+%28Poster%29%2E+Open%2DReview%2Enet%2E)
16. Yongyu Mu, Jiali Zeng, Bei Li, Xinyan Guan, Fandong Meng, Jie Zhou, Tong Xiao, and Jingbo Zhu. 2025. Dissecting long-chain-of-thought reasoning models: An empirical study. Preprint, arXiv:2506.04913. — [arXiv:2506.04913](https://arxiv.org/abs/2506.04913) [find](https://scholar.google.com/scholar?q=Yongyu+Mu%2C+Jiali+Zeng%2C+Bei+Li%2C+Xinyan+Guan%2C+Fandong+Meng%2C+Jie+Zhou%2C+Tong+Xiao%2C+and+Jingbo+Zhu%2E+2025%2E+Dissecting+long%2Dchain%2Dof%2Dthought+reasoning+models%3A+An+empirical+study%2E+Preprint%2C+arXiv%3A2506%2E04913%2E+)
17. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R. Bowman. 2023. [GPQA: A graduate-level google-proof q&a benchmark](https://md.research.elamur.ai/free/en/papers/rein-2023-gpqa-graduate-level-google-proof/). CoRR, abs/2311.12022.
18. 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](https://md.research.elamur.ai/free/en/papers/sheng-2024-hybridflow-flexible-efficient-rlhf-framework/). In EuroSys, pages 1279–1297. ACM.
19. Huang, and Junyang Lin. 2025b. [Beyond the 80/20 rule: High-entropy minority tokens drive effective reinforcement learning for LLM reasoning](https://md.research.elamur.ai/free/en/papers/wang-2025-80-20-rule-high-entropy/). CoRR, abs/2506.01939.
20. Yuhao Wang, Ruiyang Ren, Yucheng Wang, Jing Liu, Xin Zhao, Hua Wu, and Haifeng Wang. 2026. Bee-rag: Balanced entropy engineering for retrieval-augmented generation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 33737–33745. [find](https://scholar.google.com/scholar?q=Yuhao+Wang%2C+Ruiyang+Ren%2C+Yucheng+Wang%2C+Jing+Liu%2C+Xin+Zhao%2C+Hua+Wu%2C+and+Haifeng+Wang%2E+2026%2E+Bee%2Drag%3A+Balanced+entropy+engineering+for+retrieval%2Daugmented+generation%2E+In+Proceedings+of+the+AAAI+Conferen)
21. Yuhao Wang, Ruiyang Ren, Yucheng Wang, Wayne Xin Zhao, Jing Liu, Hua Wu, and Haifeng Wang. 2025c. [Reinforced informativeness optimization for long-form retrieval-augmented generation](https://md.research.elamur.ai/free/en/papers/wang-2026-reinforced-informativeness-optimization-long-form/). CoRR, abs/2505.20825.
22. Yuhao Wang, Ruiyang Ren, Yucheng Wang, Wayne Xin Zhao, Jing Liu, Hua Wu, and Haifeng Wang. 2025d. Unveiling knowledge utilization mechanisms in llm-based retrieval-augmented generation. In Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval… [find](https://scholar.google.com/scholar?q=Yuhao+Wang%2C+Ruiyang+Ren%2C+Yucheng+Wang%2C+Wayne+Xin+Zhao%2C+Jing+Liu%2C+Hua+Wu%2C+and+Haifeng+Wang%2E+2025d%2E+Unveiling+knowledge+utilization+mechanisms+in+llm%2Dbased+retrieval%2Daugmented+generation%2E+In+Proceedings)
23. Zhiheng Xi, Xin Guo, Yang Nan, Enyu Zhou, Junrui Shen, Wenxiang Chen, Jiaqi Liu, Jixuan Huang, Zhihao Zhang, Honglin Guo, Xun Deng, Zhikai Lei, Miao Zheng, Guoteng Wang, Shuo Zhang, Peng Sun, Rui Zheng, Hang Yan, Tao Gui, Qi Zhang, and Xuanjing Huang. 2025. Bapo: Stabilizing off-policy reinforcement… — [arXiv:2510.18927](https://arxiv.org/abs/2510.18927) [find](https://scholar.google.com/scholar?q=Zhiheng+Xi%2C+Xin+Guo%2C+Yang+Nan%2C+Enyu+Zhou%2C+Junrui+Shen%2C+Wenxiang+Chen%2C+Jiaqi+Liu%2C+Jixuan+Huang%2C+Zhihao+Zhang%2C+Honglin+Guo%2C+Xun+Deng%2C+Zhikai+Lei%2C+Miao+Zheng%2C+Guoteng+Wang%2C+Shuo+Zhang%2C+Peng+Sun%2C+Rui+Zhen)
24. Yuchen Yan, Liang Jiang, Jin Jiang, Shuaicheng Li, Zujie Wen, Zhiqiang Zhang, Jun Zhou, Jian Shao, Yueting Zhuang, and Yongliang Shen. 2026. Inftythink+: Effective and efficient infinite-horizon reasoning via reinforcement learning. CoRR, abs/2602.06960. — [arXiv:2602.06960](https://arxiv.org/abs/2602.06960) [find](https://scholar.google.com/scholar?q=Yuchen+Yan%2C+Liang+Jiang%2C+Jin+Jiang%2C+Shuaicheng+Li%2C+Zujie+Wen%2C+Zhiqiang+Zhang%2C+Jun+Zhou%2C+Jian+Shao%2C+Yueting+Zhuang%2C+and+Yongliang+Shen%2E+2026%2E+Inftythink%2B%3A+Effective+and+efficient+infinite%2Dhorizon+reaso)
25. Jiakang Wang, Runze Liu, Lei Lin, Wenping Hu, Xiu Li, Fuzheng Zhang, Guorui Zhou, and Kun Gai. 2025a. Aspo: Asymmetric importance sampling policy optimization. Preprint, arXiv:2510.06062. — [arXiv:2510.06062](https://arxiv.org/abs/2510.06062) [find](https://scholar.google.com/scholar?q=Jiakang+Wang%2C+Runze+Liu%2C+Lei+Lin%2C+Wenping+Hu%2C+Xiu+Li%2C+Fuzheng+Zhang%2C+Guorui+Zhou%2C+and+Kun+Gai%2E+2025a%2E+Aspo%3A+Asymmetric+importance+sampling+policy+optimization%2E+Preprint%2C+arXiv%3A2510%2E06062%2E+%E2%80%94+%5BarXiv%3A251)
26. 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, Jian Yang, Jianhong Tu, Jianwei Zhang, Jian Yang, Jiaxi Yang, Jingren Zhou, Jun… — [Qwen3 Technical Report](https://md.research.elamur.ai/free/en/papers/yang-2025-qwen3-technical-report/)
27. Shenzhi Wang, Le Yu, Chang Gao, Chujie Zheng, Shixuan Liu, Rui Lu, Kai Dang, Xionghui Chen, Jianxin Yang, Zhenru Zhang, Yuqiong Liu, An Yang, Andrew Zhao, Yang Yue, Shiji Song, Bowen Yu, Gao An Yang, Beichen Zhang, Binyuan Hui, Bofei Gao, Bowen Yu, Chengpeng Li, Dayiheng Liu, Jianhong Tu, Jingren Zh… — [Unlocking General Long Chain-of-Thought Reasoning Capabilities of Large Language Models via Representation Engineering](https://md.research.elamur.ai/free/en/papers/tang-2025-unlocking-general-long-chain-thought/)
28. Xinyu Tang, Zhenduo Zhang, Yurou Liu, Xin Zhao, zujie wen, Zhiqiang Zhang, and JUN ZHOU. 2026. Towards high data efficiency in reinforcement learning with verifiable reward. In The Fourteenth International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Xinyu+Tang%2C+Zhenduo+Zhang%2C+Yurou+Liu%2C+Xin+Zhao%2C+zujie+wen%2C+Zhiqiang+Zhang%2C+and+JUN+ZHOU%2E+2026%2E+Towards+high+data+efficiency+in+reinforcement+learning+with+verifiable+reward%2E+In+The+Fourteenth+Internat)
29. Ren, and Zhenru Zhang. 2024. [Qwen2.5-math technical report: Toward mathematical expert model via self-improvement](https://md.research.elamur.ai/free/en/papers/yang-2024-qwen2-5-math-technical-report/). CoRR, abs/2409.12122.
30. Qiying Yu, Zheng Zhang, Ruofei Zhu, Yufeng Yuan, Xiaochen Zuo, Yu Yue, Tiantian Fan, Gaohong Liu, Lingjun Liu, Xin Liu, Haibin Lin, Zhiqi Lin, Bole Ma, Guangming Sheng, Yuxuan Tong, Chi Zhang, Mofan Zhang, Wang Zhang, Hang Zhu, Jinhua Zhu, Jiaze Chen, Jiangjie Chen, Chengyi Wang, Hongli Yu, Weinan D… — [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://md.research.elamur.ai/free/en/papers/yu-2025-dapo-open-source-llm-reinforcement/)
31. Yang Yue, Zhiqi Chen, Rui Lu, Andrew Zhao, Zhaokai Wang, Yang Yue, Shiji Song, and Gao Huang. 2025a. [Does reinforcement learning really incentivize reasoning capacity in llms beyond the base model?](https://md.research.elamur.ai/free/en/papers/yue-2025-reinforcement-learning-really-incentivize-reasoning/) CoRR, abs/2504.13837.
32. Yu Yue, Yufeng Yuan, Qiying Yu, Xiaochen Zuo, Ruofei Zhu, Wenyuan Xu, Jiaze Chen, Cheng-Xiang Wang, Tiantian Fan, Zhengyin Du, Xiangpeng Wei, Xiangyu Yu, Gaohong Liu, Juncai Liu, Lingjun Liu, Haibin Lin, Zhiqi Lin, Bole Ma, Chi Zhang, Mofan Zhang, Wang Zhang, Hang Zhu, Ru Zhang, Xin Liu, Mingxuan Wa… [find](https://scholar.google.com/scholar?q=Yu+Yue%2C+Yufeng+Yuan%2C+Qiying+Yu%2C+Xiaochen+Zuo%2C+Ruofei+Zhu%2C+Wenyuan+Xu%2C+Jiaze+Chen%2C+Cheng%2DXiang+Wang%2C+Tiantian+Fan%2C+Zhengyin+Du%2C+Xiangpeng+Wei%2C+Xiangyu+Yu%2C+Gaohong+Liu%2C+Juncai+Liu%2C+Lingjun+Liu%2C+Haibin+L)
33. 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. 2023a. [A survey of large language models](https://md.research.elamur.ai/free/en/papers/zhao-2023-survey-large-language-models/). CoRR, abs/2303.18223.
34. Yanli Zhao, Andrew Gu, Rohan Varma, Liang Luo, Chien-Chin Huang, Min Xu, Less Wright, Hamid Shojanazeri, Myle Ott, Sam Shleifer, Alban Desmaison, Can Balioglu, Pritam Damania, Bernard Nguyen, Geeta Chauhan, Yuchen Hao, Ajit Mathews, and Shen Li. 2023b. Pytorch FSDP: experiences on scaling fully shar… [find](https://scholar.google.com/scholar?q=Yanli+Zhao%2C+Andrew+Gu%2C+Rohan+Varma%2C+Liang+Luo%2C+Chien%2DChin+Huang%2C+Min+Xu%2C+Less+Wright%2C+Hamid+Shojanazeri%2C+Myle+Ott%2C+Sam+Shleifer%2C+Alban+Desmaison%2C+Can+Balioglu%2C+Pritam+Damania%2C+Bernard+Nguyen%2C+Geeta+Ch)
35. Yuzhong Zhao, Yue Liu, Junpeng Liu, Jingye Chen, Xun Wu, Yaru Hao, Tengchao Lv, Shaohan Huang, Lei Cui, Qixiang Ye, Fang Wan, and Furu Wei. 2025. Geometric-mean policy optimization. CoRR, abs/2507.20673. — [arXiv:2507.20673](https://arxiv.org/abs/2507.20673) [find](https://scholar.google.com/scholar?q=Yuzhong+Zhao%2C+Yue+Liu%2C+Junpeng+Liu%2C+Jingye+Chen%2C+Xun+Wu%2C+Yaru+Hao%2C+Tengchao+Lv%2C+Shaohan+Huang%2C+Lei+Cui%2C+Qixiang+Ye%2C+Fang+Wan%2C+and+Furu+Wei%2E+2025%2E+Geometric%2Dmean+policy+optimization%2E+CoRR%2C+abs%2F2507%2E206)
36. Chujie Zheng, Shixuan Liu, Mingze Li, Xiong-Hui Chen, Bowen Yu, Chang Gao, Kai Dang, Yuqiong Liu, Rui Men, An Yang, Jingren Zhou, and Junyang Lin. 2025. [Group sequence policy optimization](https://md.research.elamur.ai/free/en/papers/zheng-2025-group-sequence-policy-optimization/). CoRR, abs/2507.18071.
37. Xinyu Zhu, Mengzhou Xia, Zhepei Wei, Wei-Lin Chen, Danqi Chen, and Yu Meng. 2025. The surprising effectiveness of negative reinforcement in LLM reasoning. CoRR, abs/2506.01347. [find](https://scholar.google.com/scholar?q=Xinyu+Zhu%2C+Mengzhou+Xia%2C+Zhepei+Wei%2C+Wei%2DLin+Chen%2C+Danqi+Chen%2C+and+Yu+Meng%2E+2025%2E+The+surprising+effectiveness+of+negative+reinforcement+in+LLM+reasoning%2E+CoRR%2C+abs%2F2506%2E01347%2E)
38. Yu-Liang Zhan, Zhong-Yi Lu, Hao Sun, and Ze-Feng Gao. 2024. Over-parameterized student model via tensor decomposition boosted knowledge distillation. In NeurIPS. [find](https://scholar.google.com/scholar?q=Yu%2DLiang+Zhan%2C+Zhong%2DYi+Lu%2C+Hao+Sun%2C+and+Ze%2DFeng+Gao%2E+2024%2E+Over%2Dparameterized+student+model+via+tensor+decomposition+boosted+knowledge+distillation%2E+In+NeurIPS%2E)
39. Yu-Liang Zhan, Xinyu Tang, Han Wan, Jian Li, Jirong Wen, and Hao Sun. 2026. L2v-cot: Cross-modal transfer of chain-of-thought reasoning via latent intervention. In AAAI, pages 12358–12366. AAAI Press. [find](https://scholar.google.com/scholar?q=Yu%2DLiang+Zhan%2C+Xinyu+Tang%2C+Han+Wan%2C+Jian+Li%2C+Jirong+Wen%2C+and+Hao+Sun%2E+2026%2E+L2v%2Dcot%3A+Cross%2Dmodal+transfer+of+chain%2Dof%2Dthought+reasoning+via+latent+intervention%2E+In+AAAI%2C+pages+12358%E2%80%9312366%2E+AAAI+Press%2E)
40. Kaiyan Zhang, Yuxin Zuo, Bingxiang He, Youbang Sun, Runze Liu, Che Jiang, Yuchen Fan, Kai Tian, Guoli Jia, Pengfei Li, Yu Fu, Xingtai Lv, Yuchen Zhang, Sihang Zeng, Shang Qu, Haozhan Li, Shijie Wang, Yuru Wang, Xinwei Long, Fangfu Liu, Xiang Xu, Jiaze Ma, Xuekai Zhu, Ermo Hua, Yihao Liu, Zonglin Li,… — [arXiv:2509.08827](https://arxiv.org/abs/2509.08827) [find](https://scholar.google.com/scholar?q=Kaiyan+Zhang%2C+Yuxin+Zuo%2C+Bingxiang+He%2C+Youbang+Sun%2C+Runze+Liu%2C+Che+Jiang%2C+Yuchen+Fan%2C+Kai+Tian%2C+Guoli+Jia%2C+Pengfei+Li%2C+Yu+Fu%2C+Xingtai+Lv%2C+Yuchen+Zhang%2C+Sihang+Zeng%2C+Shang+Qu%2C+Haozhan+Li%2C+Shijie+Wang%2C+)
41. Shengyu Zhang, Linfeng Dong, Xiaoya Li, Sen Zhang, Xiaofei Sun, Shuhe Wang, Jiwei Li, Runyi Hu, Tianwei Zhang, Fei Wu, and Guoyin Wang. 2023. [Instruction tuning for large language models: A survey](https://md.research.elamur.ai/free/en/papers/zhang-2023-instruction-tuning-large-language-models/). CoRR, abs/2308.10792.
42. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, Yifan Du, [find](https://scholar.google.com/scholar?q=Wayne+Xin+Zhao%2C+Kun+Zhou%2C+Junyi+Li%2C+Tianyi+Tang%2C+Xiaolei+Wang%2C+Yupeng+Hou%2C+Yingqian+Min%2C+Beichen+Zhang%2C+Junjie+Zhang%2C+Zican+Dong%2C+Yifan+Du%2C)

*Updated 31 October 2026*