---
title: "Reflect, Rewrite, Repeat — paper by Wang et al. (2026)"
description: "The authors propose R³ (\"reflect, rewrite, repeat\"), which needs no reward model and no policy rollouts. 45 references."
authors: ["Mengdie Flora Wang", "Haochen Xie", "Mun Young Kim", "Baishali Chaudhury", "Meghana Ashok", "Suren Gunturu", "Sungmin Hong", "Jae Oh Woo"]
year: 2026
venue: "Findings of the Association for Computational Linguistics: EACL 2026"
source: "https://aclanthology.org/2026.findings-eacl.69/"
doi: "10.18653/v1/2026.findings-eacl.69"
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/wang-2026-reflect-rewrite-repeat-how-simple/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/wang-2026-reflect-rewrite-repeat-how-simple/)

[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/41/) / Reflect, Rewrite, Repeat: How Simple Arithmetic Enables Advanced Reasoning in Small Language Models

# Reflect, Rewrite, Repeat: How Simple Arithmetic Enables Advanced Reasoning in Small Language Models

**Authors:** Mengdie Flora Wang, Haochen Xie, Mun Young Kim, Baishali Chaudhury, Meghana Ashok, Suren Gunturu, Sungmin Hong and Jae Oh Woo · **Year:** 2026 · **Venue:** [Findings of the Association for Computational Linguistics: EACL 2026](https://md.research.elamur.ai/free/en/venues/findings/) ([Findings of the Association for Computational Linguistics 2026](https://md.research.elamur.ai/free/en/venues/findings/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.findings-eacl.69/) · [PDF](https://aclanthology.org/2026.findings-eacl.69.pdf) · [DOI 10.18653/v1/2026.findings-eacl.69](https://doi.org/10.18653/v1/2026.findings-eacl.69) · [OpenAlex](https://openalex.org/W7140128505)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
The paper addresses a practical obstacle: improving reasoning in language models usually depends on reinforcement learning (RL), reward models and very large datasets, which smaller teams cannot afford. The authors propose R³ ("reflect, rewrite, repeat"), which needs no reward model and no policy rollouts. A small student model answers questions with chain-of-thought. A response counts as correct only if both the final answer and the tag format are right. When it fails, a stronger teacher model diagnoses the errors and rewrites the response, repeating until the rewrite is correct. The student's original output and the corrected version then form a preference pair for DPO, and optional SFT afterwards consolidates what was learned. Training uses only 700 elementary arithmetic problems. Even so, Qwen 2.5 7B and Qwen 2.5 Math 7B improve on mathematical, symbolic and commonsense benchmarks, and Mistral and LLaMA models gain as well. This suggests that carefully curated, teacher-guided training can stand in for large-scale RL. As a side benefit, the pipeline produces clean reasoning traces that can be reused as SFT data.

## Key points
- About 700 basic arithmetic samples are used, roughly 100× less data than typical RL pipelines. Training runs on a single GPU in a matter of hours.
- Reported gains include GSM8K 83.1% → 88.6%, AIME'25@10 20.0% → 30.0% and LastLetterConcat 40.7% → 53.3%.
- Ablations indicate that DPO and SFT complement each other, and neither alone generalizes robustly as task complexity grows. Manual analysis suggests the self-reflective phrasing in outputs reflects genuine error fixes rather than stylistic imitation.
- In a control experiment, 350 first-order logic problems replaced arithmetic, with Amazon Nova Premier as teacher and Normalized Compression Distance (threshold 0.4) for grading. This version transferred partially but trailed the arithmetic-trained model by 1.4–17.5 points, with the largest gap on AMC'23.
- Code is released at github.com/aws-samples/sample-for-reflect-rewrite-repeat. The current design depends on a teacher model, which the authors see as a temporary constraint.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 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 (2): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works · [🧮 Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/) · 1,426 works

## Cite

Cite as: Mengdie Flora Wang, Haochen Xie, Mun Young Kim, Baishali Chaudhury, Meghana Ashok, Suren Gunturu, Sungmin Hong and Jae Oh Woo. 2026. Reflect, Rewrite, Repeat: How Simple Arithmetic Enables Advanced Reasoning in Small Language Models. Findings of the Association for Computational Linguistics: EACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{wang2026reflect,
  title = {{Reflect, Rewrite, Repeat: How Simple Arithmetic Enables Advanced Reasoning in Small Language Models}},
  author = {Mengdie Flora Wang, Haochen Xie, Mun Young Kim, Baishali Chaudhury, Meghana Ashok, Suren Gunturu, Sungmin Hong and Jae Oh Woo},
  booktitle = {Findings of the Association for Computational Linguistics: EACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-eacl.69},
  url = {https://doi.org/10.18653/v1/2026.findings-eacl.69}
}
```

- [← previous Reference-Free Schema Generation for Literature Review Tables via Multi-Faceted Rewards](https://md.research.elamur.ai/free/en/papers/saha-2026-reference-free-schema-generation-literature/)
- [next → Reflection in the Dark: Exposing and Escaping the Black Box in Reflective Prompt Optimization](https://md.research.elamur.ai/free/en/papers/liu-2026-reflection-dark-exposing-escaping-black/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/41/)

## References

45 references, 22 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Yuntao Bai, Andy Jones, Kamal Ndousse, Amanda Askell, Anna Chen, Nova DasSarma, Dawn Drain, Stanislav Fort, Deep Ganguli, Tom Henighan, and 1 others. 2022. Training a helpful and harmless assistant with reinforcement learning from human feedback. arXiv preprint arXiv:2204.05862. [find](https://scholar.google.com/scholar?q=Yuntao+Bai%2C+Andy+Jones%2C+Kamal+Ndousse%2C+Amanda+Askell%2C+Anna+Chen%2C+Nova+DasSarma%2C+Dawn+Drain%2C+Stanislav+Fort%2C+Deep+Ganguli%2C+Tom+Henighan%2C+and+1+others%2E+2022%2E+Training+a+helpful+and+harmless+assistant+wi)
2. Wenhu Chen, Xueguang Ma, Xinyi Wang, and William W Cohen. 2022. Program of thoughts prompting: Disentangling computation from reasoning for numerical reasoning tasks. arXiv preprint arXiv:2211.12588. [find](https://scholar.google.com/scholar?q=Wenhu+Chen%2C+Xueguang+Ma%2C+Xinyi+Wang%2C+and+William+W+Cohen%2E+2022%2E+Program+of+thoughts+prompting%3A+Disentangling+computation+from+reasoning+for+numerical+reasoning+tasks%2E+arXiv+preprint+arXiv%3A2211%2E12588%2E)
3. Michelene TH Chi, Miriam Bassok, Matthew W Lewis, Peter Reimann, and Robert Glaser. 1989. Self-explanations: How students study and use examples in learning to solve problems. Cognitive science, 13(2):145–182. [find](https://scholar.google.com/scholar?q=Michelene+TH+Chi%2C+Miriam+Bassok%2C+Matthew+W+Lewis%2C+Peter+Reimann%2C+and+Robert+Glaser%2E+1989%2E+Self%2Dexplanations%3A+How+students+study+and+use+examples+in+learning+to+solve+problems%2E+Cognitive+science%2C+13%282%29)
4. Tianzhe Chu, Yuexiang Zhai, Jihan Yang, Shengbang Tong, Saining Xie, Dale Schuurmans, Quoc V Le, Sergey Levine, and Yi Ma. 2025. [Sft memorizes, rl generalizes: A comparative study of foundation model post-training](https://md.research.elamur.ai/free/en/papers/chu-2025-sft-memorizes-rl-generalizes-comparative/). arXiv preprint arXiv:2501.17161.
5. 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](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/). Preprint, arXiv:2110.14168.
6. 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, and 181 others. 2025. [Deepseek-r1: Incentivizing reasoning capability in ll…](https://md.research.elamur.ai/free/en/papers/deepseek-ai-2025-deepseek-r1-incentivizing-reasoning-capability/)
7. Etash Guha, Negin Raoff, Jean Mercat, Ryan Marten, Eric Frankel, Sedrick Keh, Sachin Grover, George Smyrnis, Trung Vu, Jon Saad-Falcon, Caroline Choi, Kushal Arora, Mike Merrill, Yichuan Deng, Ashima Suvarna, Hritik Bansal, Marianna Nezhurina, Reinhard Heckel, Seewong Oh, and 7 others. 2024. Evalche… [find](https://scholar.google.com/scholar?q=Etash+Guha%2C+Negin+Raoff%2C+Jean+Mercat%2C+Ryan+Marten%2C+Eric+Frankel%2C+Sedrick+Keh%2C+Sachin+Grover%2C+George+Smyrnis%2C+Trung+Vu%2C+Jon+Saad%2DFalcon%2C+Caroline+Choi%2C+Kushal+Arora%2C+Mike+Merrill%2C+Yichuan+Deng%2C+Ashima+)
8. Simeng Han, Hailey Schoelkopf, Yilun Zhao, Zhenting Qi, Martin Riddell, Luke Benson, Lucy Sun, Ekaterina Zubova, Yujie Qiao, Matthew Burtell, David Peng, Jonathan Fan, Yixin Liu, Brian Wong, Malcolm Sailor, Ansong Ni, Linyong Nan, Jungo Kasai, Tao Yu, and 7 others. 2022. [Folio: Natural language reas…](https://md.research.elamur.ai/free/en/papers/han-2022-folio-natural-language-reasoning-first/)
9. Zhiwei He. 2023. AMC23: American mathematics competitions dataset. HuggingFace Dataset. [find](https://scholar.google.com/scholar?q=Zhiwei+He%2E+2023%2E+AMC23%3A+American+mathematics+competitions+dataset%2E+HuggingFace+Dataset%2E)
10. 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/). Preprint, arXiv:2103.03874.
11. Andreas Hochlehnert, Hardik Bhatnagar, Vishaal Udandarao, Samuel Albanie, Ameya Prabhu, and Matthias Bethge. 2025. A sober look at progress in language model reasoning: Pitfalls and paths to reproducibility. Preprint, arXiv:2504.07086. [find](https://scholar.google.com/scholar?q=Andreas+Hochlehnert%2C+Hardik+Bhatnagar%2C+Vishaal+Udandarao%2C+Samuel+Albanie%2C+Ameya+Prabhu%2C+and+Matthias+Bethge%2E+2025%2E+A+sober+look+at+progress+in+language+model+reasoning%3A+Pitfalls+and+paths+to+reproduci)
12. AIME 2024: American invitational mathematics examination dataset. Hugging-Face Dataset. [find](https://scholar.google.com/scholar?q=AIME+2024%3A+American+invitational+mathematics+examination+dataset%2E+Hugging%2DFace+Dataset%2E)
13. Amazon Artificial General Intelligence. 2024. Amazon nova premier: Technical report and model card. Amazon Technical Reports. [find](https://scholar.google.com/scholar?q=Amazon+Artificial+General+Intelligence%2E+2024%2E+Amazon+nova+premier%3A+Technical+report+and+model+card%2E+Amazon+Technical+Reports%2E)
14. Xin Lai, Zhuotao Tian, Yukang Chen, Senqiao Yang, Xiangru Peng, and Jiaya Jia. 2024. [Step-dpo: Step-wise preference optimization for long-chain reasoning of llms](https://md.research.elamur.ai/free/en/papers/lai-2024-step-dpo-step-wise-preference/). arXiv preprint arXiv:2406.18629.
15. JoonHo Lee, Jae Oh Woo, Juree Seok, Parisa Hassanzadeh, Wooseok Jang, JuYoun Son, Sima Didari, Baruch Gutow, Heng Hao, Hankyu Moon, Wenjun Hu, Yeong-Dae Kwon, Taehee Lee, and Seungjai Min. 2024. Improving instruction following in language models through proxy-based uncertainty estimation. In Proceed… [find](https://scholar.google.com/scholar?q=JoonHo+Lee%2C+Jae+Oh+Woo%2C+Juree+Seok%2C+Parisa+Hassanzadeh%2C+Wooseok+Jang%2C+JuYoun+Son%2C+Sima+Didari%2C+Baruch+Gutow%2C+Heng+Hao%2C+Hankyu+Moon%2C+Wenjun+Hu%2C+Yeong%2DDae+Kwon%2C+Taehee+Lee%2C+and+Seungjai+Min%2E+2024%2E+Impro)
16. Ming Li, Xin Chen, Xin Li, Bin Ma, and Paul MB Vitányi. 2004. The similarity metric. IEEE transactions on Information Theory, 50(12):3250–3264. [find](https://scholar.google.com/scholar?q=Ming+Li%2C+Xin+Chen%2C+Xin+Li%2C+Bin+Ma%2C+and+Paul+MB+Vit%C3%A1nyi%2E+2004%2E+The+similarity+metric%2E+IEEE+transactions+on+Information+Theory%2C+50%2812%29%3A3250%E2%80%933264%2E)
17. Hunter Lightman, Vineet Kosaraju, Yura Burda, Harri Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, and Karl Cobbe. 2023. [Let’s verify step by step](https://md.research.elamur.ai/free/en/papers/lightman-2023-let-s-verify-step-step/). arXiv preprint arXiv:2305.20050.
18. Zichen Liu, Changyu Chen, Wenjun Li, Tianyu Pang, Chao Du, and Min Lin. 2025. There may not be aha moment in r1-zero-like training — a pilot study. [https://oatllm.notion.site/oat-zero](https://oatllm.notion.site/oat-zero). Notion Blog. [↗](https://oatllm.notion.site/oat-zero)
19. Aman Madaan, Niket Tandon, Prakhar Gupta, Skyler Hallinan, Luyu Gao, Sarah Wiegreffe, Uri Alon, Nouha Dziri, Shrimai Prabhumoye, Yiming Yang, and 1 others. 2023. Self-refine: Iterative refinement with self-feedback. Advances in Neural Information Processing Systems, 36:46534–46594. [find](https://scholar.google.com/scholar?q=Aman+Madaan%2C+Niket+Tandon%2C+Prakhar+Gupta%2C+Skyler+Hallinan%2C+Luyu+Gao%2C+Sarah+Wiegreffe%2C+Uri+Alon%2C+Nouha+Dziri%2C+Shrimai+Prabhumoye%2C+Yiming+Yang%2C+and+1+others%2E+2023%2E+Self%2Drefine%3A+Iterative+refinement+with)
20. Lucie Charlotte Magister, Jonathan Mallinson, Jakub Adamek, Eric Malmi, and Aliaksei Severyn. 2022. [Teaching small language models to reason](https://md.research.elamur.ai/free/en/papers/magister-2023-teaching-small-language-models-reason/). arXiv preprint arXiv:2212.08410.
21. Niklas Muennighoff, Zitong Yang, Weijia Shi, Xiang Lisa Li, Li Fei-Fei, Hannaneh Hajishirzi, Luke Zettlemoyer, Percy Liang, Emmanuel Candès, and Tatsunori Hashimoto. 2025. [s1: Simple test-time scaling](https://md.research.elamur.ai/free/en/papers/muennighoff-2025-s1-simple-test-time-scaling/). Preprint, arXiv:2501.19393.
22. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, and 1 others. 2022. [Training language models to follow instructions with human feedback](https://md.research.elamur.ai/free/en/papers/ouyang-2022-training-language-models-follow-instructions/). Advances in neural information processing systems, 35:27730–27744.
23. Richard Yuanzhe Pang, Weizhe Yuan, Kyunghyun Cho, He He, Sainbayar Sukhbaatar, and Jason Weston. 2024. [Iterative reasoning preference optimization](https://md.research.elamur.ai/free/en/papers/pang-2024-iterative-reasoning-preference-optimization/). Preprint, arXiv:2404.19733.
24. Arkil Patel, Satwik Bhattamishra, and Navin Goyal. 2021. [Are nlp models really able to solve simple math word problems?](https://md.research.elamur.ai/free/en/papers/patel-2021-are-nlp-models-really-able/) Preprint, arXiv:2103.07191.
25. Qwen, :, 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, and 25 others. 2025. Qwen2.5 technical report. Preprint, arXiv:2412.15115. [find](https://scholar.google.com/scholar?q=Qwen%2C+%3A%2C+An+Yang%2C+Baosong+Yang%2C+Beichen+Zhang%2C+Binyuan+Hui%2C+Bo+Zheng%2C+Bowen+Yu%2C+Chengyuan+Li%2C+Dayiheng+Liu%2C+Fei+Huang%2C+Haoran+Wei%2C+Huan+Lin%2C+Jian+Yang%2C+Jianhong+Tu%2C+Jianwei+Zhang%2C+Jianxin+Yang%2C+Jiaxi+)
26. 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](https://md.research.elamur.ai/free/en/papers/rafailov-2023-direct-preference-optimization-language-model/). Advances in Neural Information Processing Systems, 36:53728– 53741.
27. Victor Sanh, Lysandre Debut, Julien Chaumond, and Thomas Wolf. 2019. Distilbert, a distilled version of bert: smaller, faster, cheaper and lighter. arXiv preprint arXiv:1910.01108. [find](https://scholar.google.com/scholar?q=Victor+Sanh%2C+Lysandre+Debut%2C+Julien+Chaumond%2C+and+Thomas+Wolf%2E+2019%2E+Distilbert%2C+a+distilled+version+of+bert%3A+smaller%2C+faster%2C+cheaper+and+lighter%2E+arXiv+preprint+arXiv%3A1910%2E01108%2E)
28. Alon Talmor, Jonathan Herzig, Nicholas Lourie, and Jonathan Berant. 2018. [Commonsenseqa: A question answering challenge targeting commonsense knowledge](https://md.research.elamur.ai/free/en/papers/talmor-2019-commonsenseqa-question-answering-challenge-targeting/). arXiv preprint arXiv:1811.00937.
29. TIGER-Lab. 2024. AIME25: A benchmark for mathematical reasoning. HuggingFace Dataset. [find](https://scholar.google.com/scholar?q=TIGER%2DLab%2E+2024%2E+AIME25%3A+A+benchmark+for+mathematical+reasoning%2E+HuggingFace+Dataset%2E)
30. Peiyi Wang, Lei Li, Zhihong Shao, Runxin Xu, Damai Dai, Yifei Li, Deli Chen, Yu Wu, and Zhifang Sui. 2024. [Math-shepherd: Verify and reinforce LLMs step-by-step without human annotations](https://md.research.elamur.ai/free/en/papers/wang-2024-math-shepherd-verify-reinforce-llms/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Paper…
31. Wenhui Wang, Furu Wei, Li Dong, Hangbo Bao, Nan Yang, and Ming Zhou. 2020. Minilm: Deep self-attention distillation for task-agnostic compression of pre-trained transformers. Advances in neural information processing systems, 33:5776–5788. [find](https://scholar.google.com/scholar?q=Wenhui+Wang%2C+Furu+Wei%2C+Li+Dong%2C+Hangbo+Bao%2C+Nan+Yang%2C+and+Ming+Zhou%2E+2020%2E+Minilm%3A+Deep+self%2Dattention+distillation+for+task%2Dagnostic+compression+of+pre%2Dtrained+transformers%2E+Advances+in+neural+inform)
32. Jae Oh Woo. 2023. Active learning in bayesian neural networks with balanced entropy learning principle. In The Eleventh International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Jae+Oh+Woo%2E+2023%2E+Active+learning+in+bayesian+neural+networks+with+balanced+entropy+learning+principle%2E+In+The+Eleventh+International+Conference+on+Learning+Representations%2E)
33. Jae Oh Woo, Mengdie Flora Wang, Rahul Ghosh, Baishali Chaudhury, and Mun Young Kim. 2025. Can LLMs reliably evaluate themselves? a probabilistic VC framework. In NeurIPS 2025 Workshop on Bridging Language, Agent, and World Models for Reasoning and Planning. [find](https://scholar.google.com/scholar?q=Jae+Oh+Woo%2C+Mengdie+Flora+Wang%2C+Rahul+Ghosh%2C+Baishali+Chaudhury%2C+and+Mun+Young+Kim%2E+2025%2E+Can+LLMs+reliably+evaluate+themselves%3F+a+probabilistic+VC+framework%2E+In+NeurIPS+2025+Workshop+on+Bridging+Lang)
34. Haotian Xu, Xing Wu, Weinong Wang, Zhongzhi Li, Da Zheng, Boyuan Chen, Yi Hu, Shijia Kang, Jiaming Ji, Yingying Zhang, and 1 others. 2025a. Redstar: Does scaling long-cot data unlock better slow-reasoning systems? arXiv preprint arXiv:2501.11284. [find](https://scholar.google.com/scholar?q=Haotian+Xu%2C+Xing+Wu%2C+Weinong+Wang%2C+Zhongzhi+Li%2C+Da+Zheng%2C+Boyuan+Chen%2C+Yi+Hu%2C+Shijia+Kang%2C+Jiaming+Ji%2C+Yingying+Zhang%2C+and+1+others%2E+2025a%2E+Redstar%3A+Does+scaling+long%2Dcot+data+unlock+better+slow%2Dreaso)
35. Huimin Xu, Xin Mao, Feng-Lin Li, Xiaobao Wu, Wang Chen, Wei Zhang, and Anh Tuan Luu. 2025b. Full-step-dpo: Self-supervised preference optimization with step-wise rewards for mathematical reasoning. arXiv preprint arXiv:2502.14356. [find](https://scholar.google.com/scholar?q=Huimin+Xu%2C+Xin+Mao%2C+Feng%2DLin+Li%2C+Xiaobao+Wu%2C+Wang+Chen%2C+Wei+Zhang%2C+and+Anh+Tuan+Luu%2E+2025b%2E+Full%2Dstep%2Ddpo%3A+Self%2Dsupervised+preference+optimization+with+step%2Dwise+rewards+for+mathematical+reasoning%2E+ar)
36. An Yang, Beichen Zhang, Binyuan Hui, Bofei Gao, Bowen Yu, Chengpeng Li, Dayiheng Liu, Jianhong Tu, Jingren Zhou, Junyang Lin, Keming Lu, Mingfeng Xue, Runji Lin, Tianyu Liu, Xingzhang 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/). P…
37. Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao. 2023. [React: Synergizing reasoning and acting in language models](https://md.research.elamur.ai/free/en/papers/yao-2022-react-synergizing-reasoning-acting-language/). In International Conference on Learning Representations (ICLR).
38. Yixin Ye, Zhen Huang, Yang Xiao, Ethan Chern, Shijie Xia, and Pengfei Liu. 2025. [Limo: Less is more for reasoning](https://md.research.elamur.ai/free/en/papers/ye-2025-limo-less-more-reasoning/). Preprint, arXiv:2502.03387.
39. 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, and 16 others. 2025. [Dapo: An open-source llm reinforcement learning sys…](https://md.research.elamur.ai/free/en/papers/yu-2025-dapo-open-source-llm-reinforcement/)
40. Yang Yue, Zhiqi Chen, Rui Lu, Andrew Zhao, Zhaokai Wang, Yang Yue, Shiji Song, and Gao Huang. 2025. [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/) Preprint, arXiv:2504.13837.
41. 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](https://md.research.elamur.ai/free/en/papers/zeng-2025-simplerl-zoo-investigating-taming-zero/). Preprint, arXiv:2503.18892.
42. Chen Zhang, Chengguang Tang, Dading Chong, Ke Shi, Guohua Tang, Feng Jiang, and Haizhou Li. 2024. TS-align: A teacher-student collaborative framework for scalable iterative finetuning of large language models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 8926–8946,… [find](https://scholar.google.com/scholar?q=Chen+Zhang%2C+Chengguang+Tang%2C+Dading+Chong%2C+Ke+Shi%2C+Guohua+Tang%2C+Feng+Jiang%2C+and+Haizhou+Li%2E+2024%2E+TS%2Dalign%3A+A+teacher%2Dstudent+collaborative+framework+for+scalable+iterative+finetuning+of+large+languag)
43. Chunting Zhou, Pengfei Liu, Puxin Xu, Srinivasan Iyer, Jiao Sun, Yuning Mao, Xuezhe Ma, Avia Efrat, Ping Yu, Lili Yu, and 1 others. 2023. Lima: Less is more for alignment. Advances in Neural Information Processing Systems, 36:55006–55021. [find](https://scholar.google.com/scholar?q=Chunting+Zhou%2C+Pengfei+Liu%2C+Puxin+Xu%2C+Srinivasan+Iyer%2C+Jiao+Sun%2C+Yuning+Mao%2C+Xuezhe+Ma%2C+Avia+Efrat%2C+Ping+Yu%2C+Lili+Yu%2C+and+1+others%2E+2023%2E+Lima%3A+Less+is+more+for+alignment%2E+Advances+in+Neural+Informati)
44. Denny Zhou, Nathanael Schärli, Le Hou, Jason Wei, Nathan Scales, Xuezhi Wang, Dale Schuurmans, Claire Cui, Olivier Bousquet, Quoc Le, and 1 others. 2022. [Least-to-most prompting enables complex reasoning in large language models](https://md.research.elamur.ai/free/en/papers/al-2022-least-most-prompting-enables-complex/). arXiv preprint arXiv:2205.10625.
45. Kun Zhou, Beichen Zhang, Jiapeng Wang, Zhipeng Chen, Xin Zhao, Jing Sha, Zhichao Sheng, Shijin Wang, and Ji-Rong Wen. 2024. Jiuzhang3.0: Efficiently improving mathematical reasoning by training small data synthesis models. In The Thirty-eighth Annual Conference on Neural Information Processing Syste… [find](https://scholar.google.com/scholar?q=Kun+Zhou%2C+Beichen+Zhang%2C+Jiapeng+Wang%2C+Zhipeng+Chen%2C+Xin+Zhao%2C+Jing+Sha%2C+Zhichao+Sheng%2C+Shijin+Wang%2C+and+Ji%2DRong+Wen%2E+2024%2E+Jiuzhang3%2E0%3A+Efficiently+improving+mathematical+reasoning+by+training+small+)

*Updated 1 November 2026*