---
title: "Thinking with Reasoning Skills: Fewer Tokens, More Accuracy"
description: "The authors propose Thinking with Reasoning Skills (TRS), a training-free, retrieval-based framework… 23 references, cited by 1 paper here."
authors: ["Guangxiang Zhao", "Qilong Shi", "Xusen Xiao", "Xiangzheng Zhang", "Tong Yang", "Lin Sun"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)"
source: "https://aclanthology.org/2026.acl-industry.154/"
doi: "10.18653/v1/2026.acl-industry.154"
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/zhao-2026-thinking-reasoning-skills-fewer-tokens/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zhao-2026-thinking-reasoning-skills-fewer-tokens/)

[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/18/) / Thinking with Reasoning Skills: Fewer Tokens, More Accuracy

# Thinking with Reasoning Skills: Fewer Tokens, More Accuracy

**Authors:** [Guangxiang Zhao](https://md.research.elamur.ai/free/en/authors/guangxiang-zhao/), [Qilong Shi](https://md.research.elamur.ai/free/en/authors/qilong-shi/), Xusen Xiao, [Xiangzheng Zhang](https://md.research.elamur.ai/free/en/authors/xiangzheng-zhang/), [Tong Yang](https://md.research.elamur.ai/free/en/authors/tong-yang/) and [Lin Sun](https://md.research.elamur.ai/free/en/authors/lin-sun/) · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)](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-industry.154/) · [PDF](https://aclanthology.org/2026.acl-industry.154.pdf) · [DOI 10.18653/v1/2026.acl-industry.154](https://doi.org/10.18653/v1/2026.acl-industry.154) · [OpenAlex](https://openalex.org/W7166797905)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · 1 in our library, of the 149,439 papers whose bibliographies we read

## Summary

Reasoning LLMs such as OpenAI o1 and DeepSeek-R1 often produce thousands of intermediate tokens per query. Much of this is repeated trial-and-error, which drives up API cost and latency. Earlier efficiency methods like Chain-of-Draft, TALE and NoWait shorten the traces, but every query still starts from nothing, and accuracy often drops on hard problems. The authors propose Thinking with Reasoning Skills (TRS), a training-free, retrieval-based framework that works with black-box APIs. Offline, a stronger summarizer model turns full solution trajectories, whether successful or failed, into short structured "skill cards" indexed by keywords. Online, the system retrieves the most relevant cards for a new question and adds them to the prompt, so the model can follow a known strategy instead of rediscovering it. On math (DeepMath) and coding benchmarks, TRS uses fewer reasoning tokens while matching or beating standard CoT and compressed-prompt baselines. The gains are largest on harder problems and for weaker models. This matters for production deployments billed per token.

## Key points

- **Skill cards:**
  - Each card follows a fixed Trigger / Do / Avoid / Check / Risk format and comes with 10–20 retrieval keywords.
  - Failed trajectories become anti-pattern → correction cards.
  - Cards are stored as key–value pairs, with the original question plus keywords as the key.
- **Retrieval:** BM25 is the default for math and hybrid sparse+dense retrieval for coding. A short gating instruction tells the model to ignore inapplicable or conflicting skills.
- **External transfer test:**
  - Skills came from an AoPS-derived library of 7,616 contest problems and were evaluated on 120 questions (AIME 2024 I/II, AIME 2025, AIME 2026, HMMT November 2025).
  - Accuracy improved in 13 of 25 model–benchmark pairs and cost dropped in 20 of 25.
  - Doubao-1.8 gained +1.88 points with 2.8% lower cost; GPT-OSS-120B gained +0.50 with 6.5% lower cost.
- **Cross-model transfer:** Skills distilled from one model help other models, though skills from the same source model usually give the best accuracy.
- **Ablations:**
  - Swapping the summarizer to GPT-5-mini (500-question DeepMath subset) gave 89.4% on Doubao and 89.6% on OSS, still below full TRS on OSS.
  - The best injection prompt depends on the model: "Short" works best for Doubao and "Draft" for OSS.

*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

- [Tsinghua University](https://md.research.elamur.ai/free/en/universities/tsinghua-university/) — University, Beijing, China · 333 works · 87,019 citations
- [Peking University](https://md.research.elamur.ai/free/en/universities/peking-university/) — University, Beijing, China · 244 works · 39,359 citations
- [University of Hong Kong](https://md.research.elamur.ai/free/en/universities/university-of-hong-kong/) — University, Pok Fu Lam, China · 89 works · 11,774 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: Guangxiang Zhao, Qilong Shi, Xusen Xiao, Xiangzheng Zhang, Tong Yang and Lin Sun. 2026. Thinking with Reasoning Skills: Fewer Tokens, More Accuracy. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track).

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

```bibtex
@inproceedings{zhao2026thinking,
  title = {{Thinking with Reasoning Skills: Fewer Tokens, More Accuracy}},
  author = {Guangxiang Zhao, Qilong Shi, Xusen Xiao, Xiangzheng Zhang, Tong Yang and Lin Sun},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-industry.154},
  url = {https://doi.org/10.18653/v1/2026.acl-industry.154}
}
```

- [← previous Thinking Long, but Short: Stable Sequential Test-Time Scaling for Large Reasoning Models](https://md.research.elamur.ai/free/en/papers/metel-2026-thinking-long-but-short-stable/)
- [next → THIVLVC: Retrieval Augmented Dependency Parsing for Latin](https://md.research.elamur.ai/free/en/papers/pommeret-2026-thivlvc-retrieval-augmented-dependency-parsing/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/18/)

## References

23 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Anthropic. 2026. Introducing claude opus 4.6. [https://www.anthropic.com/news/claude-opus-4-6](https://www.anthropic.com/news/claude-opus-4-6). Accessed: 2026-02-15. [↗](https://www.anthropic.com/news/claude-opus-4-6)
2. Xiang Chen, Lei Li, Ningyu Zhang, Xiaozhuan Liang, Shumin Deng, Chuanqi Tan, Fei Huang, Luo Si, and Huajun Chen. 2022. Retrieval-augmented prompt learning. arXiv preprint arXiv:2205.14704. — [arXiv:2205.14704](https://arxiv.org/abs/2205.14704) [find](https://scholar.google.com/scholar?q=Xiang+Chen%2C+Lei+Li%2C+Ningyu+Zhang%2C+Xiaozhuan+Liang%2C+Shumin+Deng%2C+Chuanqi+Tan%2C+Fei+Huang%2C+Luo+Si%2C+and+Huajun+Chen%2E+2022%2E+Retrieval%2Daugmented+prompt+learning%2E+arXiv+preprint+arXiv%3A2205%2E14704%2E+%E2%80%94+%5BarXiv%3A22)
3. DeepSeek-AI. 2025. [Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning](https://md.research.elamur.ai/free/en/papers/deepseek-ai-2025-deepseek-r1-incentivizing-reasoning-capability/). arXiv preprint arXiv:2501.12948.
4. Google. 2025. A new era of intelligence with gemini 3. [https://blog.google/](https://blog.google/) products-and-platforms/products/gemini/ gemini-3/. Accessed: 2026-02-15. [↗](https://blog.google/)
5. Zishan Guo, Hongyu Jin, Hongshen Tan, Xin Jiang, and Yuan-Fang Wang. 2023. Prompt-guided retrieval augmentation for non-knowledge-intensive tasks. In Findings of the Association for Computational Linguistics: ACL 2023. [find](https://scholar.google.com/scholar?q=Zishan+Guo%2C+Hongyu+Jin%2C+Hongshen+Tan%2C+Xin+Jiang%2C+and+Yuan%2DFang+Wang%2E+2023%2E+Prompt%2Dguided+retrieval+augmentation+for+non%2Dknowledge%2Dintensive+tasks%2E+In+Findings+of+the+Association+for+Computational+Ling)
6. Tingxu Han, Chunrong Fang, Shiyu Zhao, Shiqing Ma, Zhenyu Chen, and Zhenting Wang. 2024. [Token-budget-aware LLM reasoning](https://md.research.elamur.ai/free/en/papers/han-2024-token-budget-aware-llm-reasoning/). Preprint, arXiv:2412.18547.
7. Shibo Hao, Sainbayar Sukhbaatar, DiJia Su, Xian Li, Zhiting Hu, Jason Weston, and Yuandong Tian. 2024. Coconut: Reasoning in continuous latent space for efficient inference. arXiv preprint arXiv:2412.06769. [find](https://scholar.google.com/scholar?q=Shibo+Hao%2C+Sainbayar+Sukhbaatar%2C+DiJia+Su%2C+Xian+Li%2C+Zhiting+Hu%2C+Jason+Weston%2C+and+Yuandong+Tian%2E+2024%2E+Coconut%3A+Reasoning+in+continuous+latent+space+for+efficient+inference%2E+arXiv+preprint+arXiv%3A2412%2E)
8. Junhao Hu, Wenrui Huang, Weidong Wang, Zhenwen Li, Tiancheng Hu, Zhixia Liu, Xusheng Chen, Tao Xie, and Yizhou Shan. 2025. Raas: Reasoning-aware attention sparsity for efficient llm reasoning. In Findings of the Association for Computational Linguistics: ACL 2025, pages 2577–2590, Vienna, Austria. A… [find](https://scholar.google.com/scholar?q=Junhao+Hu%2C+Wenrui+Huang%2C+Weidong+Wang%2C+Zhenwen+Li%2C+Tiancheng+Hu%2C+Zhixia+Liu%2C+Xusheng+Chen%2C+Tao+Xie%2C+and+Yizhou+Shan%2E+2025%2E+Raas%3A+Reasoning%2Daware+attention+sparsity+for+efficient+llm+reasoning%2E+In+Find)
9. OpenAI. 2024. [Openai o1 system card](https://md.research.elamur.ai/free/en/papers/openai-2024-openai-o1-system-card/). [https://openai.com/index/openai-o1-system-card/](https://openai.com/index/openai-o1-system-card/). Accessed: 2026-02-15. [↗](https://openai.com/index/openai-o1-system-card/)
10. OpenAI. 2025. Gpt-5 system card. [https://openai.com/index/gpt-5-system-card/](https://openai.com/index/gpt-5-system-card/). Accessed: 2026-02-15. [↗](https://openai.com/index/gpt-5-system-card/)
11. OpenAI. 2026. Openai api pricing. [https://openai.com/api/pricing/](https://openai.com/api/pricing/). Accessed: 2026-02-15. [↗](https://openai.com/api/pricing/)
12. Noah Shinn, Federico Cassano, Edward Berman, Ashwin Gopinath, Karthik Narasimhan, and Shunyu Yao. 2023. [Reflexion: Language agents with verbal reinforcement learning](https://md.research.elamur.ai/free/en/papers/shinn-2023-reflexion-language-agents-verbal-reinforcement/). arXiv preprint arXiv:2303.11366.
13. Uptime Institute. 2025. Ai reasoning will take a toll on infrastructure footprint. [https://uptimeinstitute.com/blog/ai-reasoning-will-take-a-toll-on-infrastructure-footprint](https://uptimeinstitute.com/blog/ai-reasoning-will-take-a-toll-on-infrastructure-footprint). Accessed: 2026-02-15. [↗](https://uptimeinstitute.com/blog/ai-reasoning-will-take-a-toll-on-infrastructure-footprint)
14. Chenlong Wang, Yuanning Feng, Dongping Chen, Zhaoyang Chu, Ranjay Krishna, and Tianyi Zhou. 2025. Wait, we don’t need to “wait”! removing thinking tokens improves reasoning efficiency. arXiv preprint arXiv:2506.08343. [find](https://scholar.google.com/scholar?q=Chenlong+Wang%2C+Yuanning+Feng%2C+Dongping+Chen%2C+Zhaoyang+Chu%2C+Ranjay+Krishna%2C+and+Tianyi+Zhou%2E+2025%2E+Wait%2C+we+don%E2%80%99t+need+to+%E2%80%9Cwait%E2%80%9D%21+removing+thinking+tokens+improves+reasoning+efficiency%2E+arXiv+preprint+)
15. Guanzhi Wang, Yuqi Xie, Yunfan Jiang, Ajay Mandlekar, Chaowei Xiao, Yuke Zhu, Linxi Fan, and Anima Anandkumar. 2023. Voyager: An open-ended embodied agent with large language models. arXiv preprint arXiv:2305.16291. [find](https://scholar.google.com/scholar?q=Guanzhi+Wang%2C+Yuqi+Xie%2C+Yunfan+Jiang%2C+Ajay+Mandlekar%2C+Chaowei+Xiao%2C+Yuke+Zhu%2C+Linxi+Fan%2C+and+Anima+Anandkumar%2E+2023%2E+Voyager%3A+An+open%2Dended+embodied+agent+with+large+language+models%2E+arXiv+preprint+ar)
16. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc V. Le, Ed H. Chi, and Denny Zhou. 2022. [Self-consistency improves chain of thought reasoning in language models](https://md.research.elamur.ai/free/en/papers/wang-2022-self-consistency-improves-chain-thought/). arXiv preprint arXiv:2203.11171.
17. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed H. Chi, Quoc V. Le, and Denny Zhou. 2022. [Chain-of-thought prompting elicits reasoning in large language models](https://md.research.elamur.ai/free/en/papers/wei-2022-chain-thought-prompting-elicits-reasoning/). arXiv preprint arXiv:2201.11903.
18. Mingyuan Wu, Jize Jiang, Haozhen Zheng, Meitang Li, Zhaoheng Li, Beitong Tian, Bo Chen, Yongjoo Park, Minjia Zhang, ChengXiang Zhai, and Klara Nahrstedt. 2025. Cache-of-thought: Master-apprentice framework for cost-effective vision language model reasoning. In Proceedings of the 2025 Conference on E… [find](https://scholar.google.com/scholar?q=Mingyuan+Wu%2C+Jize+Jiang%2C+Haozhen+Zheng%2C+Meitang+Li%2C+Zhaoheng+Li%2C+Beitong+Tian%2C+Bo+Chen%2C+Yongjoo+Park%2C+Minjia+Zhang%2C+ChengXiang+Zhai%2C+and+Klara+Nahrstedt%2E+2025%2E+Cache%2Dof%2Dthought%3A+Master%2Dapprentice+fram)
19. Silei Xu, Wenhao Xie, Lingxiao Zhao, and Pengcheng He. 2025. Chain of draft: Thinking faster by writing less. Preprint, arXiv:2502.18600. [find](https://scholar.google.com/scholar?q=Silei+Xu%2C+Wenhao+Xie%2C+Lingxiao+Zhao%2C+and+Pengcheng+He%2E+2025%2E+Chain+of+draft%3A+Thinking+faster+by+writing+less%2E+Preprint%2C+arXiv%3A2502%2E18600%2E)
20. Ling Yang, Zhaochen Yu, Tianjun Zhang, Shiyi Cao, Minkai Xu, Wentao Zhang, Joseph E. Gonzalez, and Bin Cui. 2024. Buffer of thoughts: Thought-augmented reasoning with large language models. arXiv preprint arXiv:2406.04271. [find](https://scholar.google.com/scholar?q=Ling+Yang%2C+Zhaochen+Yu%2C+Tianjun+Zhang%2C+Shiyi+Cao%2C+Minkai+Xu%2C+Wentao+Zhang%2C+Joseph+E%2E+Gonzalez%2C+and+Bin+Cui%2E+2024%2E+Buffer+of+thoughts%3A+Thought%2Daugmented+reasoning+with+large+language+models%2E+arXiv+prep)
21. Ling Yang, Zhaochen Yu, Tianjun Zhang, Minkai Xu, Joseph E. Gonzalez, Bin Cui, and Shuicheng Yan. 2025. Supercorrect: Advancing small llm reasoning with thought template distillation and self-correction. In International Conference on Learning Representations (ICLR). [find](https://scholar.google.com/scholar?q=Ling+Yang%2C+Zhaochen+Yu%2C+Tianjun+Zhang%2C+Minkai+Xu%2C+Joseph+E%2E+Gonzalez%2C+Bin+Cui%2C+and+Shuicheng+Yan%2E+2025%2E+Supercorrect%3A+Advancing+small+llm+reasoning+with+thought+template+distillation+and+self%2Dcorrecti)
22. Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao. 2022. [ReAct: Synergizing reasoning and acting in language models](https://md.research.elamur.ai/free/en/papers/yao-2022-react-synergizing-reasoning-acting-language/). arXiv preprint arXiv:2210.03629.
23. Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao. 2023. Tree of thoughts: Deliberate problem solving with large language models. arXiv preprint arXiv:2305.10601. [find](https://scholar.google.com/scholar?q=Shunyu+Yao%2C+Jeffrey+Zhao%2C+Dian+Yu%2C+Nan+Du%2C+Izhak+Shafran%2C+Karthik+Narasimhan%2C+and+Yuan+Cao%2E+2023%2E+Tree+of+thoughts%3A+Deliberate+problem+solving+with+large+language+models%2E+arXiv+preprint+arXiv%3A2305%2E106)

*Updated 1 November 2026*