---
title: "DeepICD-R1: Medical Reasoning through — Röhr 2026"
description: "The paper tests whether reasoning-focused post-training, so far most successful in mathematics and programming, can carry over to clinical diagnosis…"
authors: ["Tom Röhr", "Thomas Maximilian Josef Steffek", "Roman Teucher", "Keno Bressem", "Alexei Figueroa", "Paul Grundmann", "Peter Troeger", "Felix Alexander Gers", "Alexander Löser"]
year: 2026
venue: "Proceedings of the Fifteenth Language Resources and Evaluation Conference"
source: "https://aclanthology.org/2026.lrec-1.843/"
doi: "10.63317/5ntn3gnmv9cy"
licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)"
licence_url: "https://creativecommons.org/licenses/by-nc/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/rohr-2026-deepicd-r1-medical-reasoning-through/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/rohr-2026-deepicd-r1-medical-reasoning-through/)

[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/3/) / DeepICD-R1: Medical Reasoning through Hierarchical Rewards and Unsupervised Distillation

# DeepICD-R1: Medical Reasoning through Hierarchical Rewards and Unsupervised Distillation

**Authors:** Tom Röhr, Thomas Maximilian Josef Steffek, Roman Teucher, Keno Bressem, [Alexei Figueroa](https://md.research.elamur.ai/free/en/authors/alexei-figueroa/), [Paul Grundmann](https://md.research.elamur.ai/free/en/authors/paul-grundmann/), Peter Troeger, Felix Alexander Gers, [Alexander Löser](https://md.research.elamur.ai/free/en/authors/alexander-loser/) · **Year:** 2026 · **Venue:** Proceedings of the Fifteenth Language Resources and Evaluation Conference · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.lrec-1.843/) · [PDF](https://aclanthology.org/2026.lrec-1.843.pdf) · [DOI 10.63317/5ntn3gnmv9cy](https://doi.org/10.63317/5ntn3gnmv9cy) · [OpenAlex](https://openalex.org/W7160016078)

**Citations:** 0 citations (openalex-snapshot · 2026-10-14)

## Summary
The paper tests whether reasoning-focused post-training, so far most successful in mathematics and programming, can carry over to clinical diagnosis without expensive domain-specific pretraining. The authors frame the task as predicting a patient's primary discharge diagnosis from MIMIC-IV admission notes, treated as a reinforcement learning problem over the ICD-10-CM code space. A large base model is first trained with GRPO using three rewards. One checks output format. One gives partial credit for matching code prefixes at the chapter, category and full-code levels. The third is an LLM-as-a-judge score for reasoning quality. The resulting DeepICD-R1-zero-32B, combined with external ICD coding guidance and clinical validation, is used to distill MIMIC-Reason, a dataset of about 93k reasoning traces. Smaller instruction-tuned models are fine-tuned on this data and then refined with GRPO, which yields DeepICD-R1-7B. The authors report steady gains in diagnostic accuracy and reasoning coherence. Their ablations support the value of hierarchical, verifiable rewards. The work offers a reproducible, low-compute route to specialized clinical reasoning models.

## Key points
- **MIMIC-Reason dataset:** about 93k reasoning samples (the abstract says over 90k) covering roughly 6k unique ICD-10-CM codes.
  - It is built from the MIMIC-IV hosp split.
  - Sections revealing diagnoses or treatments are removed to prevent label leakage.
  - The first annotated code serves as the target.
- **Prediction levels:** the task is evaluated at three levels of the ICD hierarchy.
  - Chapter: 21 classes.
  - Category (3 characters): 1,115 classes.
  - Full diagnosis (4–7 characters): 6,369 classes.
- **Reward design:**
  - The format reward uses regular expressions. Output must contain exactly one `<think>` block and one `<diagnosis>` block, and the diagnosis must be a valid ICD-10-CM pattern.
  - The hierarchical reward scores prefix overlap with the gold code. It adds a bonus once three characters are correct, and the bonus grows linearly with each further matching character.
  - The LLM judge comes from a different model family. It rates outputs from 1 to 5 using the note, the predicted code and auxiliary ICD information, without seeing the gold label.
- **Models and analysis:** DeepICD-R1-zero-32B acts as the GRPO-trained teacher. DeepICD-R1-7B is the student, trained with SFT followed by GRPO. Ablations cover reward components and response length. Exact scores fall in the omitted middle of the supplied text, so they are not reproduced here.
- **Planned release:** code, data pipelines, reward functions, models and the SFT dataset are to be published via GitHub (DATEXIS/DeepICD-R1), HuggingFace and PhysioNet.

*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: [Linguistics & languages](https://md.research.elamur.ai/free/en/shelves/linguistics/)

## Cite

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

```bibtex
@inproceedings{rohr2026deepicd,
  title = {{DeepICD-R1: Medical Reasoning through Hierarchical Rewards and Unsupervised Distillation}},
  author = {Röhr, Tom and Steffek, Thomas Maximilian Josef and Teucher, Roman and Bressem, Keno and Figueroa, Alexei and Grundmann, Paul and Troeger, Peter and Gers, Felix Alexander and Löser, Alexander},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  year = {2026},
  doi = {10.63317/5ntn3gnmv9cy},
  url = {https://doi.org/10.63317/5ntn3gnmv9cy}
}
```

- [← previous Decomposing Sign Language Movements: A Multi-Band Visualization Method for Articulatory Analysis](https://md.research.elamur.ai/free/en/papers/sevilla-2026-decomposing-sign-language-movements-multi/)
- [next → DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence](https://md.research.elamur.ai/free/en/papers/deepseek-2026-ai-deepseek-v4-highly-efficient/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/3/)

## References

39 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Junying Chen, Zhenyang Cai, Ke Ji, et al. 2025. Towards medical complex reasoning with LLMs through medical verifiable problems. In Findings of the Association for Computational Linguistics: ACL 2025, pages 14552–14573, Vienna, Austria. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Junying+Chen%2C+Zhenyang+Cai%2C+Ke+Ji%2C+et+al%2E+2025%2E+Towards+medical+complex+reasoning+with+LLMs+through+medical+verifiable+problems%2E+In+Findings+of+the+Association+for+Computational+Linguistics%3A+ACL+2025%2C)
2. DeepSeek-AI, Daya Guo, Dejian Yang, Haowei Zhang, et al. 2025. DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning. [find](https://scholar.google.com/scholar?q=DeepSeek%2DAI%2C+Daya+Guo%2C+Dejian+Yang%2C+Haowei+Zhang%2C+et+al%2E+2025%2E+DeepSeek%2DR1%3A+Incentivizing+Reasoning+Capability+in+LLMs+via+Reinforcement+Learning%2E) [our card](https://md.research.elamur.ai/free/en/papers/deepseek-2025-ai-deepseek-r1-incentivizing-reasoning-capability/)
3. Alexei Figueroa, Jens-Michalis Papaioannou, Conor Fallon, Alexandra Bekiaridou, Keno Bressem, Stavros Zanos, Felix Gers, Wolfgang Nejdl, and Alexander Löser. 2024. Boosting Long-Tail Data Classification with Sparse Prototypical Networks. In Machine Learning and Knowledge Discovery in Databases. [find](https://scholar.google.com/scholar?q=Alexei+Figueroa%2C+Jens%2DMichalis+Papaioannou%2C+Conor+Fallon%2C+Alexandra+Bekiaridou%2C+Keno+Bressem%2C+Stavros+Zanos%2C+Felix+Gers%2C+Wolfgang+Nejdl%2C+and+Alexander+L%C3%B6ser%2E+2024%2E+Boosting+Long%2DTail+Data+Classificati)
4. Research Track: European Conference, ECML PKDD 2024, Vilnius, Lithuania, September 9–13, 2024, Proceedings, Part VII, pages 434–449, Berlin, Heidelberg. Springer-Verlag. [find](https://scholar.google.com/scholar?q=Research+Track%3A+European+Conference%2C+ECML+PKDD+2024%2C+Vilnius%2C+Lithuania%2C+September+9%E2%80%9313%2C+2024%2C+Proceedings%2C+Part+VII%2C+pages+434%E2%80%93449%2C+Berlin%2C+Heidelberg%2E+Springer%2DVerlag%2E)
5. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, et al. 2024. The llama 3 herd of models. [find](https://scholar.google.com/scholar?q=Aaron+Grattafiori%2C+Abhimanyu+Dubey%2C+Abhinav+Jauhri%2C+et+al%2E+2024%2E+The+llama+3+herd+of+models%2E)
6. Paul Grundmann, Dennis Fast, Jan Frick, Thomas Steffek, Felix Gers, Wolfgang Nejdl, and Alexander Löser. 2025. CliniBench: A Clinical Outcome Prediction Benchmark for Generative and Encoder-Based Language Models. [find](https://scholar.google.com/scholar?q=Paul+Grundmann%2C+Dennis+Fast%2C+Jan+Frick%2C+Thomas+Steffek%2C+Felix+Gers%2C+Wolfgang+Nejdl%2C+and+Alexander+L%C3%B6ser%2E+2025%2E+CliniBench%3A+A+Clinical+Outcome+Prediction+Benchmark+for+Generative+and+Encoder%2DBased+Lang)
7. Michael Hassid, Gabriel Synnaeve, Yossi Adi, and Roy Schwartz. 2025. Don’t Overthink it. Preferring Shorter Thinking Chains for Improved LLM Reasoning. [find](https://scholar.google.com/scholar?q=Michael+Hassid%2C+Gabriel+Synnaeve%2C+Yossi+Adi%2C+and+Roy+Schwartz%2E+2025%2E+Don%E2%80%99t+Overthink+it%2E+Preferring+Shorter+Thinking+Chains+for+Improved+LLM+Reasoning%2E)
8. J.A. Hirsch, G. Nicola, G. McGinty, R.W. Liu, R.M. Barr, M.D. Chittle, and L. Manchikanti. 2016. Icd-10: History and context. American Journal of Neuroradiology, 37(4):596–599. [find](https://scholar.google.com/scholar?q=J%2EA%2E+Hirsch%2C+G%2E+Nicola%2C+G%2E+McGinty%2C+R%2EW%2E+Liu%2C+R%2EM%2E+Barr%2C+M%2ED%2E+Chittle%2C+and+L%2E+Manchikanti%2E+2016%2E+Icd%2D10%3A+History+and+context%2E+American+Journal+of+Neuroradiology%2C+37%284%29%3A596%E2%80%93599%2E)
9. Kerui Huang, Shuhan Liu, Xing Hu, Tongtong Xu, Lingfeng Bao, and Xin Xia. 2025. Reasoning efficiently through adaptive chain-of-thought compression: A self-optimizing framework. [find](https://scholar.google.com/scholar?q=Kerui+Huang%2C+Shuhan+Liu%2C+Xing+Hu%2C+Tongtong+Xu%2C+Lingfeng+Bao%2C+and+Xin+Xia%2E+2025%2E+Reasoning+efficiently+through+adaptive+chain%2Dof%2Dthought+compression%3A+A+self%2Doptimizing+framework%2E)
10. Alistair E. W. Johnson, Lucas Bulgarelli, Lu Shen, Alvin Gayles, Ayad Shammout, Steven Horng, Tom J. Pollard, Sicheng Hao, Benjamin Moody, Brian Gow, Li-wei H. Lehman, Leo A. Celi, and Roger G. Mark. 2023. Mimic-iv, a freely accessible electronic health record dataset. Scientific Data, 10(1):1. [find](https://scholar.google.com/scholar?q=Alistair+E%2E+W%2E+Johnson%2C+Lucas+Bulgarelli%2C+Lu+Shen%2C+Alvin+Gayles%2C+Ayad+Shammout%2C+Steven+Horng%2C+Tom+J%2E+Pollard%2C+Sicheng+Hao%2C+Benjamin+Moody%2C+Brian+Gow%2C+Li%2Dwei+H%2E+Lehman%2C+Leo+A%2E+Celi%2C+and+Roger+G%2E+Mark%2E+) [our card](https://md.research.elamur.ai/free/en/papers/johnson-2023-mimic-iv-freely-accessible-electronic/)
11. Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph E. Gonzalez, Hao Zhang, and Ion Stoica. 2023. Efficient memory management for large language model serving with pagedattention. [find](https://scholar.google.com/scholar?q=Woosuk+Kwon%2C+Zhuohan+Li%2C+Siyuan+Zhuang%2C+Ying+Sheng%2C+Lianmin+Zheng%2C+Cody+Hao+Yu%2C+Joseph+E%2E+Gonzalez%2C+Hao+Zhang%2C+and+Ion+Stoica%2E+2023%2E+Efficient+memory+management+for+large+language+model+serving+with+p)
12. Ilya Loshchilov and Frank Hutter. 2017. Fixing weight decay regularization in adam. CoRR, abs/1711.05101. [find](https://scholar.google.com/scholar?q=Ilya+Loshchilov+and+Frank+Hutter%2E+2017%2E+Fixing+weight+decay+regularization+in+adam%2E+CoRR%2C+abs%2F1711%2E05101%2E)
13. 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. [find](https://scholar.google.com/scholar?q=Niklas+Muennighoff%2C+Zitong+Yang%2C+Weijia+Shi%2C+Xiang+Lisa+Li%2C+Li+Fei%2DFei%2C+Hannaneh+Hajishirzi%2C+Luke+Zettlemoyer%2C+Percy+Liang%2C+Emmanuel+Cand%C3%A8s%2C+and+Tatsunori+Hashimoto%2E+2025%2E+s1%3A+Simple+test%2Dtime+scaling) [our card](https://md.research.elamur.ai/free/en/papers/muennighoff-2025-s1-simple-test-time-scaling/)
14. James Mullenbach, Sarah Wiegreffe, Jon Duke, Jimeng Sun, and Jacob Eisenstein. 2018. Explainable Prediction of Medical Codes from Clinical Text. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1… [find](https://scholar.google.com/scholar?q=James+Mullenbach%2C+Sarah+Wiegreffe%2C+Jon+Duke%2C+Jimeng+Sun%2C+and+Jacob+Eisenstein%2E+2018%2E+Explainable+Prediction+of+Medical+Codes+from+Clinical+Text%2E+In+Proceedings+of+the+2018+Conference+of+the+North+Amer) [our card](https://md.research.elamur.ai/free/en/papers/mullenbach-2018-explainable-prediction-medical-codes-clinical/)
15. Long Ouyang, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul Christiano, Jan Leike, and Ryan Lowe. 2022. Train… [find](https://scholar.google.com/scholar?q=Long+Ouyang%2C+Jeff+Wu%2C+Xu+Jiang%2C+Diogo+Almeida%2C+Carroll+L%2E+Wainwright%2C+Pamela+Mishkin%2C+Chong+Zhang%2C+Sandhini+Agarwal%2C+Katarina+Slama%2C+Alex+Ray%2C+John+Schulman%2C+Jacob+Hilton%2C+Fraser+Kelton%2C+Luke+Miller%2C+)
16. Van L. Parsons. 2017. Stratified Sampling, pages 1–11. John Wiley & Sons, Ltd. [find](https://scholar.google.com/scholar?q=Van+L%2E+Parsons%2E+2017%2E+Stratified+Sampling%2C+pages+1%E2%80%9311%2E+John+Wiley+%26+Sons%2C+Ltd%2E)
17. John Pearn. 2022. Herbert french (1875-1951) and his differential diagnosis a “work of reference unique in medical literature”. J. Med. Biogr., 30(2):131–135. [find](https://scholar.google.com/scholar?q=John+Pearn%2E+2022%2E+Herbert+french+%281875%2D1951%29+and+his+differential+diagnosis+a+%E2%80%9Cwork+of+reference+unique+in+medical+literature%E2%80%9D%2E+J%2E+Med%2E+Biogr%2E%2C+30%282%29%3A131%E2%80%93135%2E)
18. Qwen, An Yang, Baosong Yang, Beichen Zhang, et al. 2025. Qwen2.5 technical report. [find](https://scholar.google.com/scholar?q=Qwen%2C+An+Yang%2C+Baosong+Yang%2C+Beichen+Zhang%2C+et+al%2E+2025%2E+Qwen2%2E5+technical+report%2E)
19. Tom Röhr, Soumyadeep Roy, Fares Al Mohamad, Jens-Michalis Papaioannou, Wolfgang Nejdl, Felix Gers, and Alexander Löser. 2025. “where does it hurt?” – dataset and study on physician intent trajectories in doctor patient dialogues. In Frontiers in Artificial Intelligence and Applications, Frontiers in… [find](https://scholar.google.com/scholar?q=Tom+R%C3%B6hr%2C+Soumyadeep+Roy%2C+Fares+Al+Mohamad%2C+Jens%2DMichalis+Papaioannou%2C+Wolfgang+Nejdl%2C+Felix+Gers%2C+and+Alexander+L%C3%B6ser%2E+2025%2E+%E2%80%9Cwhere+does+it+hurt%3F%E2%80%9D+%E2%80%93+dataset+and+study+on+physician+intent+trajectories)
20. John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. 2017. Proximal Policy Optimization Algorithms. [find](https://scholar.google.com/scholar?q=John+Schulman%2C+Filip+Wolski%2C+Prafulla+Dhariwal%2C+Alec+Radford%2C+and+Oleg+Klimov%2E+2017%2E+Proximal+Policy+Optimization+Algorithms%2E)
21. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, Y. K. Li, Y. Wu, and Daya Guo. 2024a. DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models. [find](https://scholar.google.com/scholar?q=Zhihong+Shao%2C+Peiyi+Wang%2C+Qihao+Zhu%2C+Runxin+Xu%2C+Junxiao+Song%2C+Xiao+Bi%2C+Haowei+Zhang%2C+Mingchuan+Zhang%2C+Y%2E+K%2E+Li%2C+Y%2E+Wu%2C+and+Daya+Guo%2E+2024a%2E+DeepSeekMath%3A+Pushing+the+Limits+of+Mathematical+Reasoning+i) [our card](https://md.research.elamur.ai/free/en/papers/shao-2024-deepseekmath-pushing-limits-mathematical-reasoning/)
22. Zhihong Shao, Peiyi Wang, Qihao Zhu, et al. 2024b. Deepseekmath: Pushing the limits of mathematical reasoning in open language models. [find](https://scholar.google.com/scholar?q=Zhihong+Shao%2C+Peiyi+Wang%2C+Qihao+Zhu%2C+et+al%2E+2024b%2E+Deepseekmath%3A+Pushing+the+limits+of+mathematical+reasoning+in+open+language+models%2E) [our card](https://md.research.elamur.ai/free/en/papers/shao-2024-deepseekmath-pushing-limits-mathematical-reasoning/)
23. Guangming Sheng, Chi Zhang, Zilingfeng Ye, Xibin Wu, Wang Zhang, Ru Zhang, Yanghua Peng, Haibin Lin, and Chuan Wu. 2025. Hybridflow: A flexible and efficient rlhf framework. In Proceedings of the Twentieth European Conference on Computer Systems, EuroSys ’25, page 1279–1297, New York, NY, USA. Assoc… [find](https://scholar.google.com/scholar?q=Guangming+Sheng%2C+Chi+Zhang%2C+Zilingfeng+Ye%2C+Xibin+Wu%2C+Wang+Zhang%2C+Ru+Zhang%2C+Yanghua+Peng%2C+Haibin+Lin%2C+and+Chuan+Wu%2E+2025%2E+Hybridflow%3A+A+flexible+and+efficient+rlhf+framework%2E+In+Proceedings+of+the+Twen) [our card](https://md.research.elamur.ai/free/en/papers/sheng-2024-hybridflow-flexible-efficient-rlhf-framework/)
24. Karan Singhal, Shekoofeh Azizi, Tao Tu, S. Sara Mahdavi, Jason Wei, Hyung Won Chung, Nathan Scales, Ajay Tanwani, Heather Cole-Lewis, Stephen Pfohl, Perry Payne, Martin Seneviratne, Paul Gamble, Chris Kelly, Abubakr Babiker, Nathanael Schärli, Aakanksha Chowdhery, Philip Mansfield, Dina Demner-Fushm… [find](https://scholar.google.com/scholar?q=Karan+Singhal%2C+Shekoofeh+Azizi%2C+Tao+Tu%2C+S%2E+Sara+Mahdavi%2C+Jason+Wei%2C+Hyung+Won+Chung%2C+Nathan+Scales%2C+Ajay+Tanwani%2C+Heather+Cole%2DLewis%2C+Stephen+Pfohl%2C+Perry+Payne%2C+Martin+Seneviratne%2C+Paul+Gamble%2C+Chris) [our card](https://md.research.elamur.ai/free/en/papers/singhal-2022-large-language-models-encode-clinical/)
25. Karan Singhal, Tao Tu, Juraj Gottweis, Rory Sayres, Ellery Wulczyn, Mohamed Amin, Le Hou, Kevin Clark, Stephen R. Pfohl, Heather Cole-Lewis, Darlene Neal, Qazi Mamunur Rashid, Mike Schaekermann, Amy Wang, Dev Dash, Jonathan H. Chen, Nigam H. Shah, Sami Lachgar, Philip Andrew Mansfield, Sushant Praka… [find](https://scholar.google.com/scholar?q=Karan+Singhal%2C+Tao+Tu%2C+Juraj+Gottweis%2C+Rory+Sayres%2C+Ellery+Wulczyn%2C+Mohamed+Amin%2C+Le+Hou%2C+Kevin+Clark%2C+Stephen+R%2E+Pfohl%2C+Heather+Cole%2DLewis%2C+Darlene+Neal%2C+Qazi+Mamunur+Rashid%2C+Mike+Schaekermann%2C+Amy+W)
26. Betty van Aken, Jens-Michalis Papaioannou, Marcel Naik, Georgios Eleftheriadis, Wolfgang Nejdl, Felix Gers, and Alexander Loeser. 2022. This Patient Looks Like That Patient: Prototypical Networks for Interpretable Diagnosis Prediction from Clinical Text. In Proceedings of the 2nd Conference of the A… [find](https://scholar.google.com/scholar?q=Betty+van+Aken%2C+Jens%2DMichalis+Papaioannou%2C+Marcel+Naik%2C+Georgios+Eleftheriadis%2C+Wolfgang+Nejdl%2C+Felix+Gers%2C+and+Alexander+Loeser%2E+2022%2E+This+Patient+Looks+Like+That+Patient%3A+Prototypical+Networks+for+)
27. Bowen Wang, Jiuyang Chang, Yiming Qian, Guoxin Chen, Junhao Chen, Zhouqiang Jiang, Jiahao Zhang, Yuta Nakashima, and Hajime Nagahara. 2024. Direct: Diagnostic reasoning for clinical notes via large language models. In Advances in Neural Information Processing Systems, volume 37, pages 74999–75011. C… [find](https://scholar.google.com/scholar?q=Bowen+Wang%2C+Jiuyang+Chang%2C+Yiming+Qian%2C+Guoxin+Chen%2C+Junhao+Chen%2C+Zhouqiang+Jiang%2C+Jiahao+Zhang%2C+Yuta+Nakashima%2C+and+Hajime+Nagahara%2E+2024%2E+Direct%3A+Diagnostic+reasoning+for+clinical+notes+via+large+la)
28. 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. In Proceedings of the 36th International Conference on Neural Information Processing Systems, NIPS ’22, p… [find](https://scholar.google.com/scholar?q=Jason+Wei%2C+Xuezhi+Wang%2C+Dale+Schuurmans%2C+Maarten+Bosma%2C+Brian+Ichter%2C+Fei+Xia%2C+Ed+H%2E+Chi%2C+Quoc+V%2E+Le%2C+and+Denny+Zhou%2E+2022%2E+Chain%2Dof%2Dthought+prompting+elicits+reasoning+in+large+language+models%2E+In+Pr) [our card](https://md.research.elamur.ai/free/en/papers/wei-2022-chain-thought-prompting-elicits-reasoning/)
29. Juncheng Wu, Wenlong Deng, Xingxuan Li, Sheng Liu, Taomian Mi, Yifan Peng, Ziyang Xu, Yi Liu, Hyunjin Cho, Chang-In Choi, Yihan Cao, Hui Ren, Xiang Li, Xiaoxiao Li, and Yuyin Zhou. 2025. Medreason: Eliciting factual medical reasoning steps in llms via knowledge graphs. [find](https://scholar.google.com/scholar?q=Juncheng+Wu%2C+Wenlong+Deng%2C+Xingxuan+Li%2C+Sheng+Liu%2C+Taomian+Mi%2C+Yifan+Peng%2C+Ziyang+Xu%2C+Yi+Liu%2C+Hyunjin+Cho%2C+Chang%2DIn+Choi%2C+Yihan+Cao%2C+Hui+Ren%2C+Xiang+Li%2C+Xiaoxiao+Li%2C+and+Yuyin+Zhou%2E+2025%2E+Medreason%3A+El)
30. Yixin Ye, Zhen Huang, Yang Xiao, Ethan Chern, Shijie Xia, and Pengfei Liu. 2025. Limo: Less is more for reasoning. [find](https://scholar.google.com/scholar?q=Yixin+Ye%2C+Zhen+Huang%2C+Yang+Xiao%2C+Ethan+Chern%2C+Shijie+Xia%2C+and+Pengfei+Liu%2E+2025%2E+Limo%3A+Less+is+more+for+reasoning%2E) [our card](https://md.research.elamur.ai/free/en/papers/ye-2025-limo-less-more-reasoning/)
31. Qiying Yu, Zheng Zhang, Ruofei Zhu, Yufeng Yuan, Xiaochen Zuo, Yu Yue, Weinan Dai, Tiantian Fan, Gaohong Liu, Lingjun Liu, Xin Liu, Haibin Lin, Zhiqi Lin, Bole Ma, Guangming Sheng, Yuxuan Tong, Chi Zhang, Mofan Zhang, Wang Zhang, Hang Zhu, Jinhua Zhu, Jiaze Chen, Jiangjie Chen, Chengyi Wang, Hongli… [find](https://scholar.google.com/scholar?q=Qiying+Yu%2C+Zheng+Zhang%2C+Ruofei+Zhu%2C+Yufeng+Yuan%2C+Xiaochen+Zuo%2C+Yu+Yue%2C+Weinan+Dai%2C+Tiantian+Fan%2C+Gaohong+Liu%2C+Lingjun+Liu%2C+Xin+Liu%2C+Haibin+Lin%2C+Zhiqi+Lin%2C+Bole+Ma%2C+Guangming+Sheng%2C+Yuxuan+Tong%2C+Chi+Zh) [our card](https://md.research.elamur.ai/free/en/papers/yu-2025-dapo-open-source-llm-reinforcement/)
32. Jixiao Zhang and Chunsheng Zuo. 2025. GRPO-LEAD: A Difficulty-Aware Reinforcement Learning Approach for Concise Mathematical Reasoning in Language Models. [find](https://scholar.google.com/scholar?q=Jixiao+Zhang+and+Chunsheng+Zuo%2E+2025%2E+GRPO%2DLEAD%3A+A+Difficulty%2DAware+Reinforcement+Learning+Approach+for+Concise+Mathematical+Reasoning+in+Language+Models%2E)
33. Chunting Zhou, Pengfei Liu, Puxin Xu, Srini Iyer, Jiao Sun, Yuning Mao, Xuezhe Ma, Avia Efrat, Ping Yu, Lili Yu, Susan Zhang, Gargi Ghosh, Mike Lewis, Luke Zettlemoyer, and Omer Levy. 2023. Lima: less is more for alignment. In Proceedings of the 37th International Conference on Neural Information Pr… [find](https://scholar.google.com/scholar?q=Chunting+Zhou%2C+Pengfei+Liu%2C+Puxin+Xu%2C+Srini+Iyer%2C+Jiao+Sun%2C+Yuning+Mao%2C+Xuezhe+Ma%2C+Avia+Efrat%2C+Ping+Yu%2C+Lili+Yu%2C+Susan+Zhang%2C+Gargi+Ghosh%2C+Mike+Lewis%2C+Luke+Zettlemoyer%2C+and+Omer+Levy%2E+2023%2E+Lima%3A+less)
34. Language Resource References A. L. Goldberger, L. A. Amaral, L. Glass, J. M. Hausdorff, P. C. Ivanov, R. G. Mark, J. E. Mietus, G. B. Moody, C. K. Peng, and H. E. Stanley. 2000. PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals. Circulati… [find](https://scholar.google.com/scholar?q=Language+Resource+References+A%2E+L%2E+Goldberger%2C+L%2E+A%2E+Amaral%2C+L%2E+Glass%2C+J%2E+M%2E+Hausdorff%2C+P%2E+C%2E+Ivanov%2C+R%2E+G%2E+Mark%2C+J%2E+E%2E+Mietus%2C+G%2E+B%2E+Moody%2C+C%2E+K%2E+Peng%2C+and+H%2E+E%2E+Stanley%2E+2000%2E+PhysioBank%2C+PhysioTool)
35. Alistair Johnson, Lucas Bulgarelli, Tom Pollard, Brian Gow, Benjamin Moody, Steven Horng, Leo Anthony Celi, and Roger Mark. 2024. MIMIC-IV. [find](https://scholar.google.com/scholar?q=Alistair+Johnson%2C+Lucas+Bulgarelli%2C+Tom+Pollard%2C+Brian+Gow%2C+Benjamin+Moody%2C+Steven+Horng%2C+Leo+Anthony+Celi%2C+and+Roger+Mark%2E+2024%2E+MIMIC%2DIV%2E)
36. Alistair Johnson, Tom Pollard, Steven Horng, Leo Anthony Celi, and Roger Mark. 2023. MIMIC-IV-Note: Deidentified free-text clinical notes. [find](https://scholar.google.com/scholar?q=Alistair+Johnson%2C+Tom+Pollard%2C+Steven+Horng%2C+Leo+Anthony+Celi%2C+and+Roger+Mark%2E+2023%2E+MIMIC%2DIV%2DNote%3A+Deidentified+free%2Dtext+clinical+notes%2E)
37. Tom Röhr, Alexei Figueroa, Jens-Michalis Papaioannou, Conor Fallon, Keno Bressem, Wolfgang Nejdl, and Alexander Löser. 2024. Revisiting clinical outcome prediction for MIMIC-IV. In Proceedings of the 6th Clinical Natural Language Processing Workshop, pages 208–217, Mexico City, Mexico. Association f… [find](https://scholar.google.com/scholar?q=Tom+R%C3%B6hr%2C+Alexei+Figueroa%2C+Jens%2DMichalis+Papaioannou%2C+Conor+Fallon%2C+Keno+Bressem%2C+Wolfgang+Nejdl%2C+and+Alexander+L%C3%B6ser%2E+2024%2E+Revisiting+clinical+outcome+prediction+for+MIMIC%2DIV%2E+In+Proceedings+of+the+)
38. Betty van Aken, Jens-Michalis Papaioannou, Manuel Mayrdorfer, Klemens Budde, Felix Gers, and Alexander Loeser. 2021. Clinical outcome prediction from admission notes using self-supervised knowledge integration. In Proceedings of the 16th Conference of the European Chapter of the Association for Comp… [find](https://scholar.google.com/scholar?q=Betty+van+Aken%2C+Jens%2DMichalis+Papaioannou%2C+Manuel+Mayrdorfer%2C+Klemens+Budde%2C+Felix+Gers%2C+and+Alexander+Loeser%2E+2021%2E+Clinical+outcome+prediction+from+admission+notes+using+self%2Dsupervised+knowledge+in)
39. Leandro von Werra, Younes Belkada, Lewis Tunstall, Edward Beeching, Tristan Thrush, Nathan Lambert, Shengyi Huang, Kashif Rasul, and Quentin Gallouédec. 2020. Trl: Transformer reinforcement learning. [https://github.com/huggingface/trl](https://github.com/huggingface/trl). [↗](https://github.com/huggingface/trl)