---
title: "Does Reinforcement Learning Really Incentivize — Yue 2025"
description: "Reinforcement Learning with Verifiable Rewards (RLVR) is widely credited with giving LLMs new reasoning skills in math and code."
authors: ["Yang Yue", "Zhiqi Chen", "Rui Lu", "Andrew Zhao", "Zhaokai Wang", "Yang Yue", "Shiji Song", "Gao Huang"]
year: 2025
venue: "arXiv:2504.13837 · NeurIPS 2025 Oral; ICML 2025 AI4MATH workshop best paper"
source: "https://arxiv.org/abs/2504.13837"
doi: "10.48550/arxiv.2504.13837"
arxiv: "2504.13837"
licence: "CC BY 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/yue-2025-reinforcement-learning-really-incentivize-reasoning/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/yue-2025-reinforcement-learning-really-incentivize-reasoning/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/) / Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?

# Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?

**Authors:** Yang Yue, Zhiqi Chen, [Rui Lu](https://md.research.elamur.ai/free/en/authors/rui-lu/), Andrew Zhao, [Zhaokai Wang](https://md.research.elamur.ai/free/en/authors/zhaokai-wang/), Yang Yue, Shiji Song, Gao Huang · **Year:** 2025 · **Venue:** arXiv:2504.13837 · NeurIPS 2025 Oral; ICML 2025 AI4MATH workshop best paper · **Licence:** [CC BY 4.0 (arXiv)](http://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2504.13837) · [PDF](https://arxiv.org/pdf/2504.13837) · [DOI 10.48550/arxiv.2504.13837](https://doi.org/10.48550/arxiv.2504.13837) · [arXiv 2504.13837](https://arxiv.org/abs/2504.13837) · [OpenAlex](https://openalex.org/W6966810623)

**Citations:** 3 citations (OpenAlex, 2026-10-14) · [193](https://md.research.elamur.ai/free/en/papers/yue-2025-reinforcement-learning-really-incentivize-reasoning/cited-by/) in our library, of the 140884 papers whose bibliographies we read

## Summary
Reinforcement Learning with Verifiable Rewards (RLVR) is widely credited with giving LLMs new reasoning skills in math and code. This paper tests that belief. Instead of average accuracy, the authors use pass@k at large k (up to 256–1024 samples) as a proxy for a model's reasoning boundary: the set of problems it can solve at all. They compare base models with their RLVR-trained versions across several model families and sizes (Qwen2.5, LLaMA-3.1), six RL algorithms, and math, coding and visual-reasoning benchmarks. RLVR models win when k is small, but base models catch up and overtake them as k grows. Accuracy-distribution and perplexity analyses show that the reasoning paths of RLVR models already appear in the base model's own outputs. So RLVR makes sampling more efficient but does not add new capability, and the solvable range can even shrink during training. Distillation from a stronger teacher, by contrast, does widen this range. For anyone fine-tuning reasoning models, the takeaway is that current RLVR mostly redistributes probability mass. Better exploration, process-level rewards, curated data at scale and multi-turn agentic RL are proposed as the way forward.

## Key points
- **Metric:** pass@k counts a problem as solved if any of k samples is correct. Base models beat RLVR models at large k on AIME24, MATH500, Minerva, OlympiadBench, GSM8K and AMC23, for Qwen2.5-7B/14B/32B and LLaMA-3.1-8B.
- **Narrowing during training:** with GRPO on Omni-MATH-Train for Qwen2.5-7B (checkpoints at steps 150, 300 and 450), pass@1 goes up while pass@256 goes down.
- **Sampling efficiency gap (ΔSE):** this is the RL model's pass@1 minus the base model's pass@256. It stays large and nearly the same across six algorithms, including PPO, GRPO and Reinforce++, so none of them comes close to the base model's potential.
- **Perplexity/coverage:** the base model assigns high likelihood to the outputs of RLVR models, which suggests RLVR introduces no new reasoning patterns.
- **Distillation differs:** models distilled from a stronger teacher reach above the base model's pass@k ceiling. Project page: limit-of-RLVR.github.io.

*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: [Mathematics & statistics](https://md.research.elamur.ai/free/en/shelves/mathematics/)

Topics: [Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · [Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/)

## Cite

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

```bibtex
@misc{yue2025does,
  title = {{Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?}},
  author = {Yue, Yang and Chen, Zhiqi and Lu, Rui and Zhao, Andrew and Wang, Zhaokai and Yue, Yang and Song, Shiji and Huang, Gao},
  year = {2025},
  doi = {10.48550/arxiv.2504.13837},
  eprint = {2504.13837},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2504.13837}
}
```

- [← previous LIMO: Less is More for Reasoning](https://md.research.elamur.ai/free/en/papers/ye-2025-limo-less-more-reasoning/)
- [next → InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://md.research.elamur.ai/free/en/papers/zhu-2025-internvl3-exploring-advanced-training/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/)

## References

63 references, 33 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Achiam, J., Adler, S., Agarwal, S., Ahmad, L., Akkaya, I., Aleman, F. L., Almeida, D., Altenschmidt, J., Altman, S., Anadkat, S., et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774, 2023. 2 [find](https://scholar.google.com/scholar?q=Achiam%2C+J%2E%2C+Adler%2C+S%2E%2C+Agarwal%2C+S%2E%2C+Ahmad%2C+L%2E%2C+Akkaya%2C+I%2E%2C+Aleman%2C+F%2E+L%2E%2C+Almeida%2C+D%2E%2C+Altenschmidt%2C+J%2E%2C+Altman%2C+S%2E%2C+Anadkat%2C+S%2E%2C+et+al%2E+Gpt%2D4+technical+report%2E+arXiv+preprint+arXiv%3A2303%2E08774%2C+2023%2E+) [our card](https://md.research.elamur.ai/free/en/papers/openai-2023-gpt-4-technical-report/)
2. Ahmadian, A., Cremer, C., Gallé, M., Fadaee, M., Kreutzer, J., Pietquin, O., Üstun, A., and Hooker, S. Back to basics: Revisiting reinforce style optimization for learning from human feedback in llms. ACL, 2024. 9, 19 [find](https://scholar.google.com/scholar?q=Ahmadian%2C+A%2E%2C+Cremer%2C+C%2E%2C+Gall%C3%A9%2C+M%2E%2C+Fadaee%2C+M%2E%2C+Kreutzer%2C+J%2E%2C+Pietquin%2C+O%2E%2C+%C3%9Cstun%2C+A%2E%2C+and+Hooker%2C+S%2E+Back+to+basics%3A+Revisiting+reinforce+style+optimization+for+learning+from+human+feedback+in+llms%2E)
3. Bai, S., Chen, K., Liu, X., Wang, J., Ge, W., Song, S., Dang, K., Wang, P., Wang, S., Tang, J., et al. Qwen2.5-vl technical report. arXiv preprint arXiv:2502.13923, 2025. 7 [find](https://scholar.google.com/scholar?q=Bai%2C+S%2E%2C+Chen%2C+K%2E%2C+Liu%2C+X%2E%2C+Wang%2C+J%2E%2C+Ge%2C+W%2E%2C+Song%2C+S%2E%2C+Dang%2C+K%2E%2C+Wang%2C+P%2E%2C+Wang%2C+S%2E%2C+Tang%2C+J%2E%2C+et+al%2E+Qwen2%2E5%2Dvl+technical+report%2E+arXiv+preprint+arXiv%3A2502%2E13923%2C+2025%2E+7) [our card](https://md.research.elamur.ai/free/en/papers/bai-2025-qwen2-5-vl-technical/)
4. Brown, B., Juravsky, J., Ehrlich, R., Clark, R., Le, Q. V., Ré, C., and Mirhoseini, A. Large language monkeys: Scaling inference compute with repeated sampling. arXiv preprint arXiv:2407.21787, 2024. 2, 6 [find](https://scholar.google.com/scholar?q=Brown%2C+B%2E%2C+Juravsky%2C+J%2E%2C+Ehrlich%2C+R%2E%2C+Clark%2C+R%2E%2C+Le%2C+Q%2E+V%2E%2C+R%C3%A9%2C+C%2E%2C+and+Mirhoseini%2C+A%2E+Large+language+monkeys%3A+Scaling+inference+compute+with+repeated+sampling%2E+arXiv+preprint+arXiv%3A2407%2E21787%2C+2024%2E+) [our card](https://md.research.elamur.ai/free/en/papers/brown-2024-large-language-monkeys-scaling-inference/)
5. Chen, L., Li, L., Zhao, H., Song, Y., and Vinci. R1-v: Reinforcing super generalization ability in vision-language models with less than $3. [https://github.com/Deep-Agent/R1-V](https://github.com/Deep-Agent/R1-V), 2025a. Accessed: 2025-02-02. 7, 19 [↗](https://github.com/Deep-Agent/R1-V)
6. Chen, M., Tworek, J., Jun, H., Yuan, Q., Pinto, H. P. D. O., Kaplan, J., Edwards, H., Burda, Y., Joseph, N., Brockman, G., et al. Evaluating large language models trained on code. arXiv preprint arXiv:2107.03374, 2021. 5, 19 [find](https://scholar.google.com/scholar?q=Chen%2C+M%2E%2C+Tworek%2C+J%2E%2C+Jun%2C+H%2E%2C+Yuan%2C+Q%2E%2C+Pinto%2C+H%2E+P%2E+D%2E+O%2E%2C+Kaplan%2C+J%2E%2C+Edwards%2C+H%2E%2C+Burda%2C+Y%2E%2C+Joseph%2C+N%2E%2C+Brockman%2C+G%2E%2C+et+al%2E+Evaluating+large+language+models+trained+on+code%2E+arXiv+preprint+arXiv) [our card](https://md.research.elamur.ai/free/en/papers/chen-2021-evaluating-large-language-models-trained/)
7. Chen, Y., Yang, Z., Liu, Z., Lee, C., Xu, P., Shoeybi, M., Catanzaro, B., and Ping, W. Acereason-nemotron: Advancing math and code reasoning through reinforcement learning. arXiv preprint arXiv:2505.16400, 2025b. 12 [find](https://scholar.google.com/scholar?q=Chen%2C+Y%2E%2C+Yang%2C+Z%2E%2C+Liu%2C+Z%2E%2C+Lee%2C+C%2E%2C+Xu%2C+P%2E%2C+Shoeybi%2C+M%2E%2C+Catanzaro%2C+B%2E%2C+and+Ping%2C+W%2E+Acereason%2Dnemotron%3A+Advancing+math+and+code+reasoning+through+reinforcement+learning%2E+arXiv+preprint+arXiv%3A2505%2E1)
8. Cobbe, K., Kosaraju, V., Bavarian, M., Chen, M., Jun, H., Kaiser, L., Plappert, M., Tworek, J., Hilton, J., Nakano, R., Hesse, C., and Schulman, J. Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168, 2021. 5, 6 [find](https://scholar.google.com/scholar?q=Cobbe%2C+K%2E%2C+Kosaraju%2C+V%2E%2C+Bavarian%2C+M%2E%2C+Chen%2C+M%2E%2C+Jun%2C+H%2E%2C+Kaiser%2C+L%2E%2C+Plappert%2C+M%2E%2C+Tworek%2C+J%2E%2C+Hilton%2C+J%2E%2C+Nakano%2C+R%2E%2C+Hesse%2C+C%2E%2C+and+Schulman%2C+J%2E+Training+verifiers+to+solve+math+word+problems%2E+arXi) [our card](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/)
9. Dang, X., Baek, C., Kolter, J. Z., and Raghunathan, A. Assessing diversity collapse in reasoning. In Scaling Self-Improving Foundation Models without Human Supervision, 2025. URL [https://openreview.net/forum?id=AMiKsHLjQh](https://openreview.net/forum?id=AMiKsHLjQh). 12, 19 [↗](https://openreview.net/forum?id=AMiKsHLjQh)
10. Gao, B., Song, F., Yang, Z., Cai, Z., Miao, Y., Dong, Q., Li, L., Ma, C., Chen, L., Xu, R., Tang, Z., Wang, B., Zan, D., Quan, S., Zhang, G., Sha, L., Zhang, Y., Ren, X., Liu, T., and Chang, B. Omni-math: A universal olympiad level mathematic benchmark for large language models, 2025. 9 [find](https://scholar.google.com/scholar?q=Gao%2C+B%2E%2C+Song%2C+F%2E%2C+Yang%2C+Z%2E%2C+Cai%2C+Z%2E%2C+Miao%2C+Y%2E%2C+Dong%2C+Q%2E%2C+Li%2C+L%2E%2C+Ma%2C+C%2E%2C+Chen%2C+L%2E%2C+Xu%2C+R%2E%2C+Tang%2C+Z%2E%2C+Wang%2C+B%2E%2C+Zan%2C+D%2E%2C+Quan%2C+S%2E%2C+Zhang%2C+G%2E%2C+Sha%2C+L%2E%2C+Zhang%2C+Y%2E%2C+Ren%2C+X%2E%2C+Liu%2C+T%2E%2C+and+Chang%2C+B%2E+Omni%2Dm) [our card](https://md.research.elamur.ai/free/en/papers/gao-2024-omni-math-universal-olympiad-level/)
11. Grattafiori, A., Dubey, A., Jauhri, A., Pandey, A., Kadian, A., Al-Dahle, A., Letman, A., Mathur, A., Schelten, A., Vaughan, A., et al. The llama 3 herd of models. arXiv preprint arXiv:2407.21783, 2024. 2, 5, 6 [find](https://scholar.google.com/scholar?q=Grattafiori%2C+A%2E%2C+Dubey%2C+A%2E%2C+Jauhri%2C+A%2E%2C+Pandey%2C+A%2E%2C+Kadian%2C+A%2E%2C+Al%2DDahle%2C+A%2E%2C+Letman%2C+A%2E%2C+Mathur%2C+A%2E%2C+Schelten%2C+A%2E%2C+Vaughan%2C+A%2E%2C+et+al%2E+The+llama+3+herd+of+models%2E+arXiv+preprint+arXiv%3A2407%2E21783%2C+202)
12. Gulcehre, C., Paine, T. L., Srinivasan, S., Konyushkova, K., Weerts, L., Sharma, A., Siddhant, A., Ahern, A., Wang, M., Gu, C., et al. Reinforced self-training (rest) for language modeling. arXiv preprint arXiv:2308.08998, 2023. 19 [find](https://scholar.google.com/scholar?q=Gulcehre%2C+C%2E%2C+Paine%2C+T%2E+L%2E%2C+Srinivasan%2C+S%2E%2C+Konyushkova%2C+K%2E%2C+Weerts%2C+L%2E%2C+Sharma%2C+A%2E%2C+Siddhant%2C+A%2E%2C+Ahern%2C+A%2E%2C+Wang%2C+M%2E%2C+Gu%2C+C%2E%2C+et+al%2E+Reinforced+self%2Dtraining+%28rest%29+for+language+modeling%2E+arXiv+prep) [our card](https://md.research.elamur.ai/free/en/papers/gulcehre-2023-reinforced-self-training-rest-language/)
13. Guo, D., Yang, D., Zhang, H., Song, J., Zhang, R., Xu, R., Zhu, Q., Ma, S., Wang, P., Bi, X., et al. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948, 2025. 1, 2, 4, 9, 12, 19 [find](https://scholar.google.com/scholar?q=Guo%2C+D%2E%2C+Yang%2C+D%2E%2C+Zhang%2C+H%2E%2C+Song%2C+J%2E%2C+Zhang%2C+R%2E%2C+Xu%2C+R%2E%2C+Zhu%2C+Q%2E%2C+Ma%2C+S%2E%2C+Wang%2C+P%2E%2C+Bi%2C+X%2E%2C+et+al%2E+Deepseek%2Dr1%3A+Incentivizing+reasoning+capability+in+llms+via+reinforcement+learning%2E+arXiv+preprint+) [our card](https://md.research.elamur.ai/free/en/papers/deepseek-2025-ai-deepseek-r1-incentivizing-reasoning-capability/)
14. He, C., Luo, R., Bai, Y., Hu, S., Thai, Z. L., Shen, J., Hu, J., Han, X., Huang, Y., Zhang, Y., et al. Olympiadbench: A challenging benchmark for promoting agi with olympiad-level bilingual multimodal scientific problems. ACL, 2024. 6 [find](https://scholar.google.com/scholar?q=He%2C+C%2E%2C+Luo%2C+R%2E%2C+Bai%2C+Y%2E%2C+Hu%2C+S%2E%2C+Thai%2C+Z%2E+L%2E%2C+Shen%2C+J%2E%2C+Hu%2C+J%2E%2C+Han%2C+X%2E%2C+Huang%2C+Y%2E%2C+Zhang%2C+Y%2E%2C+et+al%2E+Olympiadbench%3A+A+challenging+benchmark+for+promoting+agi+with+olympiad%2Dlevel+bilingual+multimodal) [our card](https://md.research.elamur.ai/free/en/papers/he-2024-olympiadbench-challenging-benchmark-promoting-agi/)
15. Hendrycks, D., Burns, C., Kadavath, S., Arora, A., Basart, S., Tang, E., Song, D., and Steinhardt, J. Measuring mathematical problem solving with the math dataset. NeurIPS, 2021. 6 [find](https://scholar.google.com/scholar?q=Hendrycks%2C+D%2E%2C+Burns%2C+C%2E%2C+Kadavath%2C+S%2E%2C+Arora%2C+A%2E%2C+Basart%2C+S%2E%2C+Tang%2C+E%2E%2C+Song%2C+D%2E%2C+and+Steinhardt%2C+J%2E+Measuring+mathematical+problem+solving+with+the+math+dataset%2E+NeurIPS%2C+2021%2E+6) [our card](https://md.research.elamur.ai/free/en/papers/hendrycks-2021-measuring-mathematical-problem-solving-math/)
16. Holtzman, A., Buys, J., Du, L., Forbes, M., and Choi, Y. The curious case of neural text degeneration. ICLR, 2020. 2, 5 [find](https://scholar.google.com/scholar?q=Holtzman%2C+A%2E%2C+Buys%2C+J%2E%2C+Du%2C+L%2E%2C+Forbes%2C+M%2E%2C+and+Choi%2C+Y%2E+The+curious+case+of+neural+text+degeneration%2E+ICLR%2C+2020%2E+2%2C+5)
17. Hu, J. Reinforce++: A simple and efficient approach for aligning large language models. arXiv preprint arXiv:2501.03262, 2025. 9 [find](https://scholar.google.com/scholar?q=Hu%2C+J%2E+Reinforce%2B%2B%3A+A+simple+and+efficient+approach+for+aligning+large+language+models%2E+arXiv+preprint+arXiv%3A2501%2E03262%2C+2025%2E+9)
18. Huang, Y. and Yang, L. F. Winning gold at imo 2025 with a model-agnostic verification-and-refinement pipeline. arXiv preprint arXiv:2507.15855, 2025. 12 — [arXiv:2507.15855](https://arxiv.org/abs/2507.15855) [find](https://scholar.google.com/scholar?q=Huang%2C+Y%2E+and+Yang%2C+L%2E+F%2E+Winning+gold+at+imo+2025+with+a+model%2Dagnostic+verification%2Dand%2Drefinement+pipeline%2E+arXiv+preprint+arXiv%3A2507%2E15855%2C+2025%2E+12+%E2%80%94+%5BarXiv%3A2507%2E15855%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2507)
19. Jaech, A., Kalai, A., Lerer, A., Richardson, A., El-Kishky, A., Low, A., Helyar, A., Madry, A., Beutel, A., Carney, A., et al. Openai o1 system card. arXiv preprint arXiv:2412.16720, 2024. 1, 8, 19 [find](https://scholar.google.com/scholar?q=Jaech%2C+A%2E%2C+Kalai%2C+A%2E%2C+Lerer%2C+A%2E%2C+Richardson%2C+A%2E%2C+El%2DKishky%2C+A%2E%2C+Low%2C+A%2E%2C+Helyar%2C+A%2E%2C+Madry%2C+A%2E%2C+Beutel%2C+A%2E%2C+Carney%2C+A%2E%2C+et+al%2E+Openai+o1+system+card%2E+arXiv+preprint+arXiv%3A2412%2E16720%2C+2024%2E+1%2C+8%2C+19) [our card](https://md.research.elamur.ai/free/en/papers/openai-2024-openai-o1-system-card/)
20. Jain, N., Han, K., Gu, A., Li, W.-D., Yan, F., Zhang, T., Wang, S., Solar-Lezama, A., Sen, K., and Stoica, I. Livecodebench: Holistic and contamination free evaluation of large language models for code. ICLR, 2025. 7 [find](https://scholar.google.com/scholar?q=Jain%2C+N%2E%2C+Han%2C+K%2E%2C+Gu%2C+A%2E%2C+Li%2C+W%2E%2DD%2E%2C+Yan%2C+F%2E%2C+Zhang%2C+T%2E%2C+Wang%2C+S%2E%2C+Solar%2DLezama%2C+A%2E%2C+Sen%2C+K%2E%2C+and+Stoica%2C+I%2E+Livecodebench%3A+Holistic+and+contamination+free+evaluation+of+large+language+models+for+cod) [our card](https://md.research.elamur.ai/free/en/papers/jain-2024-livecodebench-holistic-contamination-free-evaluation/)
21. Lambert, N., Morrison, J., Pyatkin, V., Huang, S., Ivison, H., Brahman, F., Miranda, L. J. V., Liu, A., Dziri, N., Lyu, S., et al. Tulu 3: Pushing frontiers in open language model post-training. arXiv preprint arXiv:2411.15124, 2024. 2, 12, 19 [find](https://scholar.google.com/scholar?q=Lambert%2C+N%2E%2C+Morrison%2C+J%2E%2C+Pyatkin%2C+V%2E%2C+Huang%2C+S%2E%2C+Ivison%2C+H%2E%2C+Brahman%2C+F%2E%2C+Miranda%2C+L%2E+J%2E+V%2E%2C+Liu%2C+A%2E%2C+Dziri%2C+N%2E%2C+Lyu%2C+S%2E%2C+et+al%2E+Tulu+3%3A+Pushing+frontiers+in+open+language+model+post%2Dtraining%2E+arXiv) [our card](https://md.research.elamur.ai/free/en/papers/lambert-2024-tulu-3-pushing-frontiers-open/)
22. Lewkowycz, A., Andreassen, A., Dohan, D., Dyer, E., Michalewski, H., Ramasesh, V., Slone, A., Anil, C., Schlag, I., Gutman-Solo, T., et al. Solving quantitative reasoning problems with language models. NeurIPS, 2022. 6 [find](https://scholar.google.com/scholar?q=Lewkowycz%2C+A%2E%2C+Andreassen%2C+A%2E%2C+Dohan%2C+D%2E%2C+Dyer%2C+E%2E%2C+Michalewski%2C+H%2E%2C+Ramasesh%2C+V%2E%2C+Slone%2C+A%2E%2C+Anil%2C+C%2E%2C+Schlag%2C+I%2E%2C+Gutman%2DSolo%2C+T%2E%2C+et+al%2E+Solving+quantitative+reasoning+problems+with+language+models) [our card](https://md.research.elamur.ai/free/en/papers/lewkowycz-2022-solving-quantitative-reasoning-problems-language/)
23. Li, J., Lin, H., Lu, H., Wen, K., Yang, Z., Gao, J., Wu, Y., and Zhang, J. Questa: Expanding reasoning capacity in llms via question augmentation. arXiv preprint arXiv:2507.13266, 2025. 12 [find](https://scholar.google.com/scholar?q=Li%2C+J%2E%2C+Lin%2C+H%2E%2C+Lu%2C+H%2E%2C+Wen%2C+K%2E%2C+Yang%2C+Z%2E%2C+Gao%2C+J%2E%2C+Wu%2C+Y%2E%2C+and+Zhang%2C+J%2E+Questa%3A+Expanding+reasoning+capacity+in+llms+via+question+augmentation%2E+arXiv+preprint+arXiv%3A2507%2E13266%2C+2025%2E+12)
24. Li, Z., Xu, T., Zhang, Y., Lin, Z., Yu, Y., Sun, R., and Luo, Z.-Q. Remax: A simple, effective, and efficient reinforcement learning method for aligning large language models. ICML, 2024. 9 [find](https://scholar.google.com/scholar?q=Li%2C+Z%2E%2C+Xu%2C+T%2E%2C+Zhang%2C+Y%2E%2C+Lin%2C+Z%2E%2C+Yu%2C+Y%2E%2C+Sun%2C+R%2E%2C+and+Luo%2C+Z%2E%2DQ%2E+Remax%3A+A+simple%2C+effective%2C+and+efficient+reinforcement+learning+method+for+aligning+large+language+models%2E+ICML%2C+2024%2E+9)
25. Liu, A., Feng, B., Xue, B., Wang, B., Wu, B., Lu, C., Zhao, C., Deng, C., Zhang, C., Ruan, C., et al. Deepseek-v3 technical report. arXiv preprint arXiv:2412.19437, 2024. 5 [find](https://scholar.google.com/scholar?q=Liu%2C+A%2E%2C+Feng%2C+B%2E%2C+Xue%2C+B%2E%2C+Wang%2C+B%2E%2C+Wu%2C+B%2E%2C+Lu%2C+C%2E%2C+Zhao%2C+C%2E%2C+Deng%2C+C%2E%2C+Zhang%2C+C%2E%2C+Ruan%2C+C%2E%2C+et+al%2E+Deepseek%2Dv3+technical+report%2E+arXiv+preprint+arXiv%3A2412%2E19437%2C+2024%2E+5)
26. Liu, J. and Zhang, L. Code-r1: Reproducing r1 for code with reliable rewards. [https://github.com/](https://github.com/) ganler/code-r1, 2025. GitHub repository. 7, 19 [↗](https://github.com/)
27. Liu, J., Xia, C. S., Wang, Y., and Zhang, L. Is your code generated by chatGPT really correct? rigorous evaluation of large language models for code generation. In NeurIPS, 2023. 7 [find](https://scholar.google.com/scholar?q=Liu%2C+J%2E%2C+Xia%2C+C%2E+S%2E%2C+Wang%2C+Y%2E%2C+and+Zhang%2C+L%2E+Is+your+code+generated+by+chatGPT+really+correct%3F+rigorous+evaluation+of+large+language+models+for+code+generation%2E+In+NeurIPS%2C+2023%2E+7) [our card](https://md.research.elamur.ai/free/en/papers/liu-2023-your-code-generated-chatgpt-really/)
28. Liu, Z., Chen, C., Li, W., Pang, T., Du, C., and Lin, M. 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), 2025a. Notion Blog. 12, 19 [↗](https://oatllm.notion.site/oat-zero)
29. Liu, Z., Chen, C., Li, W., Qi, P., Pang, T., Du, C., Lee, W. S., and Lin, M. Understanding r1-zero-like training: A critical perspective. arXiv preprint arXiv:2503.20783, 2025b. 6, 9, 19 [find](https://scholar.google.com/scholar?q=Liu%2C+Z%2E%2C+Chen%2C+C%2E%2C+Li%2C+W%2E%2C+Qi%2C+P%2E%2C+Pang%2C+T%2E%2C+Du%2C+C%2E%2C+Lee%2C+W%2E+S%2E%2C+and+Lin%2C+M%2E+Understanding+r1%2Dzero%2Dlike+training%3A+A+critical+perspective%2E+arXiv+preprint+arXiv%3A2503%2E20783%2C+2025b%2E+6%2C+9%2C+19) [our card](https://md.research.elamur.ai/free/en/papers/liu-2025-understanding-r1-zero-like-training/)
30. Loshchilov, I. and Hutter, F. Decoupled weight decay regularization. In ICLR, 2017. 9 [find](https://scholar.google.com/scholar?q=Loshchilov%2C+I%2E+and+Hutter%2C+F%2E+Decoupled+weight+decay+regularization%2E+In+ICLR%2C+2017%2E+9)
31. Lu, P., Gong, R., Jiang, S., Qiu, L., Huang, S., Liang, X., and Zhu, S.-C. Inter-gps: Interpretable geometry problem solving with formal language and symbolic reasoning. In ACL, 2021. 7 [find](https://scholar.google.com/scholar?q=Lu%2C+P%2E%2C+Gong%2C+R%2E%2C+Jiang%2C+S%2E%2C+Qiu%2C+L%2E%2C+Huang%2C+S%2E%2C+Liang%2C+X%2E%2C+and+Zhu%2C+S%2E%2DC%2E+Inter%2Dgps%3A+Interpretable+geometry+problem+solving+with+formal+language+and+symbolic+reasoning%2E+In+ACL%2C+2021%2E+7) [our card](https://md.research.elamur.ai/free/en/papers/lu-2021-inter-gps-interpretable-geometry-problem/)
32. Lu, P., Bansal, H., Xia, T., Liu, J., Li, C., Hajishirzi, H., Cheng, H., Chang, K.-W., Galley, M., and Gao, J. Mathvista: Evaluating mathematical reasoning of foundation models in visual contexts. In ICLR, 2024. 7 [find](https://scholar.google.com/scholar?q=Lu%2C+P%2E%2C+Bansal%2C+H%2E%2C+Xia%2C+T%2E%2C+Liu%2C+J%2E%2C+Li%2C+C%2E%2C+Hajishirzi%2C+H%2E%2C+Cheng%2C+H%2E%2C+Chang%2C+K%2E%2DW%2E%2C+Galley%2C+M%2E%2C+and+Gao%2C+J%2E+Mathvista%3A+Evaluating+mathematical+reasoning+of+foundation+models+in+visual+contexts%2E+In+) [our card](https://md.research.elamur.ai/free/en/papers/lu-2023-mathvista-evaluating-mathematical-reasoning-foundation/)
33. Luo, M., Tan, S., Huang, R., Patel, A., Ariyak, A., Wu, Q., Shi, X., Xin, R., Cai, C., Weber, M., Zhang, C., Li, L. E., Popa, R. A., and Stoica, I. Deepcoder: A fully open-source 14b coder at o3-mini level, 2025. Notion Blog. 7 [find](https://scholar.google.com/scholar?q=Luo%2C+M%2E%2C+Tan%2C+S%2E%2C+Huang%2C+R%2E%2C+Patel%2C+A%2E%2C+Ariyak%2C+A%2E%2C+Wu%2C+Q%2E%2C+Shi%2C+X%2E%2C+Xin%2C+R%2E%2C+Cai%2C+C%2E%2C+Weber%2C+M%2E%2C+Zhang%2C+C%2E%2C+Li%2C+L%2E+E%2E%2C+Popa%2C+R%2E+A%2E%2C+and+Stoica%2C+I%2E+Deepcoder%3A+A+fully+open%2Dsource+14b+coder+at+o3%2Dmini+)
34. Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. Human-level control through deep reinforcement learning. nature, 518(7540):529–533, 2015. 2, 11 [find](https://scholar.google.com/scholar?q=Mnih%2C+V%2E%2C+Kavukcuoglu%2C+K%2E%2C+Silver%2C+D%2E%2C+Rusu%2C+A%2E+A%2E%2C+Veness%2C+J%2E%2C+Bellemare%2C+M%2E+G%2E%2C+Graves%2C+A%2E%2C+Riedmiller%2C+M%2E%2C+Fidjeland%2C+A%2E+K%2E%2C+Ostrovski%2C+G%2E%2C+et+al%2E+Human%2Dlevel+control+through+deep+reinforcement+lea)
35. Novikov, A., Vũ, N., Eisenberger, M., Dupont, E., Huang, P.-S., Wagner, A. Z., Shirobokov, S., Kozlovskii, B., Ruiz, F. J., Mehrabian, A., et al. Alphaevolve: A coding agent for scientific and algorithmic discovery. arXiv preprint arXiv:2506.13131, 2025. 12 [find](https://scholar.google.com/scholar?q=Novikov%2C+A%2E%2C+Vu%CC%83%2C+N%2E%2C+Eisenberger%2C+M%2E%2C+Dupont%2C+E%2E%2C+Huang%2C+P%2E%2DS%2E%2C+Wagner%2C+A%2E+Z%2E%2C+Shirobokov%2C+S%2E%2C+Kozlovskii%2C+B%2E%2C+Ruiz%2C+F%2E+J%2E%2C+Mehrabian%2C+A%2E%2C+et+al%2E+Alphaevolve%3A+A+coding+agent+for+scientific+and+algori) [our card](https://md.research.elamur.ai/free/en/papers/novikov-2025-alphaevolve-coding-agent-scientific-algorithmic/)
36. Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., et al. Training language models to follow instructions with human feedback. NeurIPS, 2022. 19 [find](https://scholar.google.com/scholar?q=Ouyang%2C+L%2E%2C+Wu%2C+J%2E%2C+Jiang%2C+X%2E%2C+Almeida%2C+D%2E%2C+Wainwright%2C+C%2E%2C+Mishkin%2C+P%2E%2C+Zhang%2C+C%2E%2C+Agarwal%2C+S%2E%2C+Slama%2C+K%2E%2C+Ray%2C+A%2E%2C+et+al%2E+Training+language+models+to+follow+instructions+with+human+feedback%2E+NeurIPS) [our card](https://md.research.elamur.ai/free/en/papers/ouyang-2022-training-language-models-follow-instructions/)
37. Rafailov, R., Sharma, A., Mitchell, E., Manning, C. D., Ermon, S., and Finn, C. Direct preference optimization: Your language model is secretly a reward model. NeurIPS, 2023. 19 [find](https://scholar.google.com/scholar?q=Rafailov%2C+R%2E%2C+Sharma%2C+A%2E%2C+Mitchell%2C+E%2E%2C+Manning%2C+C%2E+D%2E%2C+Ermon%2C+S%2E%2C+and+Finn%2C+C%2E+Direct+preference+optimization%3A+Your+language+model+is+secretly+a+reward+model%2E+NeurIPS%2C+2023%2E+19) [our card](https://md.research.elamur.ai/free/en/papers/rafailov-2023-direct-preference-optimization-language-model/)
38. Ramamurthy, R., Ammanabrolu, P., Brantley, K., Hessel, J., Sifa, R., Bauckhage, C., Hajishirzi, H., and Choi, Y. Is reinforcement learning (not) for natural language processing: Benchmarks, baselines, and building blocks for natural language policy optimization. In ICLR, 2023. 11 [find](https://scholar.google.com/scholar?q=Ramamurthy%2C+R%2E%2C+Ammanabrolu%2C+P%2E%2C+Brantley%2C+K%2E%2C+Hessel%2C+J%2E%2C+Sifa%2C+R%2E%2C+Bauckhage%2C+C%2E%2C+Hajishirzi%2C+H%2E%2C+and+Choi%2C+Y%2E+Is+reinforcement+learning+%28not%29+for+natural+language+processing%3A+Benchmarks%2C+baselines%2C) [our card](https://md.research.elamur.ai/free/en/papers/ramamurthy-2022-reinforcement-learning-natural-language-processing/)
39. Rastogi, A., Jiang, A. Q., Lo, A., Berrada, G., Lample, G., Rute, J., Barmentlo, J., Yadav, K., Khandelwal, K., Chandu, K. R., et al. Magistral. arXiv preprint arXiv:2506.10910, 2025. 11 — [arXiv:2506.10910](https://arxiv.org/abs/2506.10910) [find](https://scholar.google.com/scholar?q=Rastogi%2C+A%2E%2C+Jiang%2C+A%2E+Q%2E%2C+Lo%2C+A%2E%2C+Berrada%2C+G%2E%2C+Lample%2C+G%2E%2C+Rute%2C+J%2E%2C+Barmentlo%2C+J%2E%2C+Yadav%2C+K%2E%2C+Khandelwal%2C+K%2E%2C+Chandu%2C+K%2E+R%2E%2C+et+al%2E+Magistral%2E+arXiv+preprint+arXiv%3A2506%2E10910%2C+2025%2E+11+%E2%80%94+%5BarXiv%3A2506)
40. Schulman, J., Wolski, F., Dhariwal, P., Radford, A., and Klimov, O. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347, 2017. 3, 9 [find](https://scholar.google.com/scholar?q=Schulman%2C+J%2E%2C+Wolski%2C+F%2E%2C+Dhariwal%2C+P%2E%2C+Radford%2C+A%2E%2C+and+Klimov%2C+O%2E+Proximal+policy+optimization+algorithms%2E+arXiv+preprint+arXiv%3A1707%2E06347%2C+2017%2E+3%2C+9)
41. Shah, D. J., Rushton, P., Singla, S., Parmar, M., Smith, K., Vanjani, Y., Vaswani, A., Chaluvaraju, A., Hojel, A., Ma, A., et al. Rethinking reflection in pre-training. arXiv preprint arXiv:2504.04022, 2025. 12, 19 [find](https://scholar.google.com/scholar?q=Shah%2C+D%2E+J%2E%2C+Rushton%2C+P%2E%2C+Singla%2C+S%2E%2C+Parmar%2C+M%2E%2C+Smith%2C+K%2E%2C+Vanjani%2C+Y%2E%2C+Vaswani%2C+A%2E%2C+Chaluvaraju%2C+A%2E%2C+Hojel%2C+A%2E%2C+Ma%2C+A%2E%2C+et+al%2E+Rethinking+reflection+in+pre%2Dtraining%2E+arXiv+preprint+arXiv%3A2504%2E04022)
42. Shao, Z., Wang, P., Zhu, Q., Xu, R., Song, J., Bi, X., Zhang, H., Zhang, M., Li, Y., Wu, Y., et al. Deepseekmath: Pushing the limits of mathematical reasoning in open language models. arXiv preprint arXiv:2402.03300, 2024. 9, 12, 19 [find](https://scholar.google.com/scholar?q=Shao%2C+Z%2E%2C+Wang%2C+P%2E%2C+Zhu%2C+Q%2E%2C+Xu%2C+R%2E%2C+Song%2C+J%2E%2C+Bi%2C+X%2E%2C+Zhang%2C+H%2E%2C+Zhang%2C+M%2E%2C+Li%2C+Y%2E%2C+Wu%2C+Y%2E%2C+et+al%2E+Deepseekmath%3A+Pushing+the+limits+of+mathematical+reasoning+in+open+language+models%2E+arXiv+preprint+a) [our card](https://md.research.elamur.ai/free/en/papers/shao-2024-deepseekmath-pushing-limits-mathematical-reasoning/)
43. Shen, H., Zhang, Z., Zhao, K., Zhang, Q., Xu, R., and Zhao, T. Vlm-r1: A stable and generalizable r1-style large vision-language model. [https://github.com/om-ai-lab/VLM-R1](https://github.com/om-ai-lab/VLM-R1), 2025. Accessed: 2025-02-15. 7, 19 [↗](https://github.com/om-ai-lab/VLM-R1)
44. Sheng, G., Zhang, C., Ye, Z., Wu, X., Zhang, W., Zhang, R., Peng, Y., Lin, H., and Wu, C. Hybridflow: A flexible and efficient rlhf framework. arXiv preprint arXiv: 2409.19256, 2024. 9 [find](https://scholar.google.com/scholar?q=Sheng%2C+G%2E%2C+Zhang%2C+C%2E%2C+Ye%2C+Z%2E%2C+Wu%2C+X%2E%2C+Zhang%2C+W%2E%2C+Zhang%2C+R%2E%2C+Peng%2C+Y%2E%2C+Lin%2C+H%2E%2C+and+Wu%2C+C%2E+Hybridflow%3A+A+flexible+and+efficient+rlhf+framework%2E+arXiv+preprint+arXiv%3A+2409%2E19256%2C+2024%2E+9) [our card](https://md.research.elamur.ai/free/en/papers/sheng-2024-hybridflow-flexible-efficient-rlhf-framework/)
45. Silver, D. and Sutton, R. S. Welcome to the era of experience. Google AI, 2025. 12 [find](https://scholar.google.com/scholar?q=Silver%2C+D%2E+and+Sutton%2C+R%2E+S%2E+Welcome+to+the+era+of+experience%2E+Google+AI%2C+2025%2E+12)
46. Silver, D., Schrittwieser, J., Simonyan, K., Antonoglou, I., Huang, A., Guez, A., Hubert, T., Baker, L., Lai, M., Bolton, A., et al. Mastering the game of go without human knowledge. nature, 550(7676): 354–359, 2017. 2, 11 [find](https://scholar.google.com/scholar?q=Silver%2C+D%2E%2C+Schrittwieser%2C+J%2E%2C+Simonyan%2C+K%2E%2C+Antonoglou%2C+I%2E%2C+Huang%2C+A%2E%2C+Guez%2C+A%2E%2C+Hubert%2C+T%2E%2C+Baker%2C+L%2E%2C+Lai%2C+M%2E%2C+Bolton%2C+A%2E%2C+et+al%2E+Mastering+the+game+of+go+without+human+knowledge%2E+nature%2C+550%287676%29)
47. Sutton, R. S., Barto, A. G., et al. Reinforcement learning: An introduction, volume 1. MIT press Cambridge, 1998. 4 [find](https://scholar.google.com/scholar?q=Sutton%2C+R%2E+S%2E%2C+Barto%2C+A%2E+G%2E%2C+et+al%2E+Reinforcement+learning%3A+An+introduction%2C+volume+1%2E+MIT+press+Cambridge%2C+1998%2E+4)
48. Team, K., Du, A., Gao, B., Xing, B., Jiang, C., Chen, C., Li, C., Xiao, C., Du, C., Liao, C., et al. Kimi k1. 5: Scaling reinforcement learning with llms. arXiv preprint arXiv:2501.12599, 2025. 1 [find](https://scholar.google.com/scholar?q=Team%2C+K%2E%2C+Du%2C+A%2E%2C+Gao%2C+B%2E%2C+Xing%2C+B%2E%2C+Jiang%2C+C%2E%2C+Chen%2C+C%2E%2C+Li%2C+C%2E%2C+Xiao%2C+C%2E%2C+Du%2C+C%2E%2C+Liao%2C+C%2E%2C+et+al%2E+Kimi+k1%2E+5%3A+Scaling+reinforcement+learning+with+llms%2E+arXiv+preprint+arXiv%3A2501%2E12599%2C+2025%2E+1) [our card](https://md.research.elamur.ai/free/en/papers/kimi-2025-team-kimi-k1-5-scaling-reinforcement/)
49. Wang, K., Pan, J., Shi, W., Lu, Z., Ren, H., Zhou, A., Zhan, M., and Li, H. Measuring multimodal mathematical reasoning with math-vision dataset. In NeurIPS Datasets and Benchmarks Track, 2024. 7 [find](https://scholar.google.com/scholar?q=Wang%2C+K%2E%2C+Pan%2C+J%2E%2C+Shi%2C+W%2E%2C+Lu%2C+Z%2E%2C+Ren%2C+H%2E%2C+Zhou%2C+A%2E%2C+Zhan%2C+M%2E%2C+and+Li%2C+H%2E+Measuring+multimodal+mathematical+reasoning+with+math%2Dvision+dataset%2E+In+NeurIPS+Datasets+and+Benchmarks+Track%2C+2024%2E+7) [our card](https://md.research.elamur.ai/free/en/papers/wang-2024-measuring-multimodal-mathematical-reasoning-math/)
50. Wang, S., Yu, L., Gao, C., Zheng, C., Liu, S., Lu, R., Dang, K., Chen, X., Yang, J., Zhang, Z., et al. Beyond the 80/20 rule: High-entropy minority tokens drive effective reinforcement learning for llm reasoning. NeurIPS, 2025. 19 [find](https://scholar.google.com/scholar?q=Wang%2C+S%2E%2C+Yu%2C+L%2E%2C+Gao%2C+C%2E%2C+Zheng%2C+C%2E%2C+Liu%2C+S%2E%2C+Lu%2C+R%2E%2C+Dang%2C+K%2E%2C+Chen%2C+X%2E%2C+Yang%2C+J%2E%2C+Zhang%2C+Z%2E%2C+et+al%2E+Beyond+the+80%2F20+rule%3A+High%2Dentropy+minority+tokens+drive+effective+reinforcement+learning+for+ll) [our card](https://md.research.elamur.ai/free/en/papers/wang-2025-80-20-rule-high-entropy/)
51. Wang, Y., Ivison, H., Dasigi, P., Hessel, J., Khot, T., Chandu, K., Wadden, D., MacMillan, K., Smith, N. A., Beltagy, I., et al. How far can camels go? exploring the state of instruction tuning on open resources. NeurIPS, 2023. 19 [find](https://scholar.google.com/scholar?q=Wang%2C+Y%2E%2C+Ivison%2C+H%2E%2C+Dasigi%2C+P%2E%2C+Hessel%2C+J%2E%2C+Khot%2C+T%2E%2C+Chandu%2C+K%2E%2C+Wadden%2C+D%2E%2C+MacMillan%2C+K%2E%2C+Smith%2C+N%2E+A%2E%2C+Beltagy%2C+I%2E%2C+et+al%2E+How+far+can+camels+go%3F+exploring+the+state+of+instruction+tuning+on+ope) [our card](https://md.research.elamur.ai/free/en/papers/wang-2023-far-camels-go-exploring-state/)
52. Williams, R. J. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine learning, 8:229–256, 1992. 4 [find](https://scholar.google.com/scholar?q=Williams%2C+R%2E+J%2E+Simple+statistical+gradient%2Dfollowing+algorithms+for+connectionist+reinforcement+learning%2E+Machine+learning%2C+8%3A229%E2%80%93256%2C+1992%2E+4)
53. Yang, A., Yang, B., Zhang, B., Hui, B., Zheng, B., Yu, B., Li, C., Liu, D., Huang, F., Wei, H., et al. Qwen2.5 technical report. arXiv preprint arXiv:2412.15115, 2024. 5, 6 [find](https://scholar.google.com/scholar?q=Yang%2C+A%2E%2C+Yang%2C+B%2E%2C+Zhang%2C+B%2E%2C+Hui%2C+B%2E%2C+Zheng%2C+B%2E%2C+Yu%2C+B%2E%2C+Li%2C+C%2E%2C+Liu%2C+D%2E%2C+Huang%2C+F%2E%2C+Wei%2C+H%2E%2C+et+al%2E+Qwen2%2E5+technical+report%2E+arXiv+preprint+arXiv%3A2412%2E15115%2C+2024%2E+5%2C+6)
54. Yang, A., Li, A., Yang, B., Zhang, B., Hui, B., Zheng, B., Yu, B., Gao, C., Huang, C., Lv, C., et al. Qwen3 technical report. arXiv preprint arXiv:2505.09388, 2025a. 11 [find](https://scholar.google.com/scholar?q=Yang%2C+A%2E%2C+Li%2C+A%2E%2C+Yang%2C+B%2E%2C+Zhang%2C+B%2E%2C+Hui%2C+B%2E%2C+Zheng%2C+B%2E%2C+Yu%2C+B%2E%2C+Gao%2C+C%2E%2C+Huang%2C+C%2E%2C+Lv%2C+C%2E%2C+et+al%2E+Qwen3+technical+report%2E+arXiv+preprint+arXiv%3A2505%2E09388%2C+2025a%2E+11) [our card](https://md.research.elamur.ai/free/en/papers/yang-2025-qwen3-technical-report/)
55. Yang, A., Yu, B., Li, C., Liu, D., Huang, F., Huang, H., Jiang, J., Tu, J., Zhang, J., Zhou, J., et al. Qwen2.5-1m technical report. arXiv preprint arXiv:2501.15383, 2025b. 7 [find](https://scholar.google.com/scholar?q=Yang%2C+A%2E%2C+Yu%2C+B%2E%2C+Li%2C+C%2E%2C+Liu%2C+D%2E%2C+Huang%2C+F%2E%2C+Huang%2C+H%2E%2C+Jiang%2C+J%2E%2C+Tu%2C+J%2E%2C+Zhang%2C+J%2E%2C+Zhou%2C+J%2E%2C+et+al%2E+Qwen2%2E5%2D1m+technical+report%2E+arXiv+preprint+arXiv%3A2501%2E15383%2C+2025b%2E+7) [our card](https://md.research.elamur.ai/free/en/papers/yang-2025-qwen2-5-1m-technical/)
56. Yu, Q., Zhang, Z., Zhu, R., Yuan, Y., Zuo, X., Yue, Y., Fan, T., Liu, G., Liu, L., Liu, X., et al. Dapo: An open-source llm reinforcement learning system at scale. arXiv preprint arXiv:2503.14476, 2025. 6, 9, 10, 19 [find](https://scholar.google.com/scholar?q=Yu%2C+Q%2E%2C+Zhang%2C+Z%2E%2C+Zhu%2C+R%2E%2C+Yuan%2C+Y%2E%2C+Zuo%2C+X%2E%2C+Yue%2C+Y%2E%2C+Fan%2C+T%2E%2C+Liu%2C+G%2E%2C+Liu%2C+L%2E%2C+Liu%2C+X%2E%2C+et+al%2E+Dapo%3A+An+open%2Dsource+llm+reinforcement+learning+system+at+scale%2E+arXiv+preprint+arXiv%3A2503%2E14476%2C+202) [our card](https://md.research.elamur.ai/free/en/papers/yu-2025-dapo-open-source-llm-reinforcement/)
57. Yue, Y., Kang, B., Xu, Z., Huang, G., and Yan, S. Value-consistent representation learning for data-efficient reinforcement learning. In AAAI, 2023. 11 [find](https://scholar.google.com/scholar?q=Yue%2C+Y%2E%2C+Kang%2C+B%2E%2C+Xu%2C+Z%2E%2C+Huang%2C+G%2E%2C+and+Yan%2C+S%2E+Value%2Dconsistent+representation+learning+for+data%2Defficient+reinforcement+learning%2E+In+AAAI%2C+2023%2E+11)
58. Zelikman, E., Wu, Y., Mu, J., and Goodman, N. Star: Bootstrapping reasoning with reasoning. NeurIPS, 2022. 19 [find](https://scholar.google.com/scholar?q=Zelikman%2C+E%2E%2C+Wu%2C+Y%2E%2C+Mu%2C+J%2E%2C+and+Goodman%2C+N%2E+Star%3A+Bootstrapping+reasoning+with+reasoning%2E+NeurIPS%2C+2022%2E+19) [our card](https://md.research.elamur.ai/free/en/papers/zelikman-2022-star-bootstrapping-reasoning-reasoning/)
59. Zeng, W., Huang, Y., Liu, Q., Liu, W., He, K., Ma, Z., and He, J. Simplerl-zoo: Investigating and taming zero reinforcement learning for open base models in the wild. arXiv preprint arXiv:2503.18892, 2025. 6, 19 [find](https://scholar.google.com/scholar?q=Zeng%2C+W%2E%2C+Huang%2C+Y%2E%2C+Liu%2C+Q%2E%2C+Liu%2C+W%2E%2C+He%2C+K%2E%2C+Ma%2C+Z%2E%2C+and+He%2C+J%2E+Simplerl%2Dzoo%3A+Investigating+and+taming+zero+reinforcement+learning+for+open+base+models+in+the+wild%2E+arXiv+preprint+arXiv%3A2503%2E18892%2C+) [our card](https://md.research.elamur.ai/free/en/papers/zeng-2025-simplerl-zoo-investigating-taming-zero/)
60. Zhang, K., Lv, A., Li, J., Wang, Y., Wang, F., Hu, H., and Yan, R. Stephint: Multi-level stepwise hints enhance reinforcement learning to reason. arXiv preprint arXiv:2507.02841, 2025. 12 [find](https://scholar.google.com/scholar?q=Zhang%2C+K%2E%2C+Lv%2C+A%2E%2C+Li%2C+J%2E%2C+Wang%2C+Y%2E%2C+Wang%2C+F%2E%2C+Hu%2C+H%2E%2C+and+Yan%2C+R%2E+Stephint%3A+Multi%2Dlevel+stepwise+hints+enhance+reinforcement+learning+to+reason%2E+arXiv+preprint+arXiv%3A2507%2E02841%2C+2025%2E+12)
61. Zhao, A., Wu, Y., Yue, Y., Wu, T., Xu, Q., Lin, M., Wang, S., Wu, Q., Zheng, Z., and Huang, G. Absolute zero: Reinforced self-play reasoning with zero data. NeurIPS, 2025a. 19 [find](https://scholar.google.com/scholar?q=Zhao%2C+A%2E%2C+Wu%2C+Y%2E%2C+Yue%2C+Y%2E%2C+Wu%2C+T%2E%2C+Xu%2C+Q%2E%2C+Lin%2C+M%2E%2C+Wang%2C+S%2E%2C+Wu%2C+Q%2E%2C+Zheng%2C+Z%2E%2C+and+Huang%2C+G%2E+Absolute+zero%3A+Reinforced+self%2Dplay+reasoning+with+zero+data%2E+NeurIPS%2C+2025a%2E+19)
62. Zhao, R., Meterez, A., Kakade, S., Pehlevan, C., Jelassi, S., and Malach, E. Echo chamber: Rl post-training amplifies behaviors learned in pretraining. arXiv preprint arXiv:2504.07912, 2025b. 12, 19 [find](https://scholar.google.com/scholar?q=Zhao%2C+R%2E%2C+Meterez%2C+A%2E%2C+Kakade%2C+S%2E%2C+Pehlevan%2C+C%2E%2C+Jelassi%2C+S%2E%2C+and+Malach%2C+E%2E+Echo+chamber%3A+Rl+post%2Dtraining+amplifies+behaviors+learned+in+pretraining%2E+arXiv+preprint+arXiv%3A2504%2E07912%2C+2025b%2E+12%2C+19)
63. Zheng, Y., Lu, J., Wang, S., Feng, Z., Kuang, D., and Xiong, Y. Easyr1: An efficient, scalable, multi-modality rl training framework. [https://github.com/hiyouga/EasyR1](https://github.com/hiyouga/EasyR1), 2025. 7, 19 [↗](https://github.com/hiyouga/EasyR1)