Elamur Research⌕
☰
home / papers / 2026 / ReasonAny: Incorporating Reasoning Capability to Any Model via Simple and Effective Model Merging

ReasonAny: Incorporating Reasoning Capability to Any Model via Simple and Effective Model Merging

Paper · 2026 · Junyao Yang, Chen Qian, Wen Shen et al.

Summary

The paper tackles "Reasoning + X": giving domain-specialized LLMs (safety, biomedicine, finance) the long chain-of-thought skills of Large Reasoning Models. The usual route is SFT or RL on domain reasoning data, which is costly. Model merging needs no training, but the authors show that current merging methods produce a destructive performance collapse, where the merged model loses both reasoning depth and domain competence.

They measured gradient nuclear norms on attention projections. Reasoning (long-CoT) models had clearly smaller gradients than ordinary instruct models. This suggests that reasoning is stored in low-gradient parameters, whereas domain knowledge is normally linked to high-magnitude ones.

ReasonAny builds on this through Contrastive Gradient Identification. It keeps high-gradient entries of the domain task vector and low-gradient entries of the reasoning task vector. It then makes the two masks mutually exclusive and adds both onto the base model. Across the three domains, ReasonAny beats established merging baselines while largely keeping reasoning ability. This makes it an inexpensive way to add reasoning to existing specialized models.

Key points

  • Gradient analysis models: DeepSeek-R1-Distill reasoning models built on Qwen2.5-7B, Qwen2.5-14B and Llama3.1-8B. Their nuclear norms on Q/K/V/O matrices were consistently lower than those of the vanilla-CoT base models.
  • Additive test on Qwen2.5-7B: the authors added only the lowest-gradient part of the reasoning task vector back into the base model. This restored GSM8K to 83.18 at 10% sparsity and to 70.79 at smaller ratios. Adding the highest-gradient part instead caused total collapse.
  • Baselines: Linear, Task Arithmetic, Ties, DARE, FuseLLM and LED.
  • Safety setting results: ReasonAny averaged 57.46 across the reasoning and knowledge benchmarks (GSM8K, Math, AIME, HumanEval, LiveCodeBench, ARC, MMLU, GPQA). This compares with 44.63 for Task Arithmetic and 44.14 for LED; the reasoning model itself scores 64.89. ReasonAny also kept harmfulness low, with Safety-Tuned 1.29 and HarmBench 0.03.
  • Limitations stated by the authors:
    • The method assumes reasoning and domain parameters do not overlap.
    • It has only been tested for merging two models.
    • Gradient attribution makes it more expensive than plain weight averaging.

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

🎓National University of SingaporeSingapore · Singapore207 works · 39,576 citations🎓Renmin University of ChinaBeijing · China139 works · 17,752 citations🔬Shanghai Artificial Intelligence LaboratoryShanghai · China152 works · 13,164 citations🎓Tongji UniversityShanghai · China33 works · 2,993 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksComputer Vision and Pattern Recognition · 3,486 works

🧭 Topics 1Research subjects and languages the work is about.Reasoning · 1,685 works

Cite

Cite as Junyao Yang, Chen Qian, Wen Shen, Yong Liu, Jing Shao and Dongrui Liu. 2026. ReasonAny: Incorporating Reasoning Capability to Any Model via Simple and Effective Model Merging. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.2201

@inproceedings{yang2026reasonany,
  title = {{ReasonAny: Incorporating Reasoning Capability to Any Model via Simple and Effective Model Merging}},
  author = {Junyao Yang, Chen Qian, Wen Shen, Yong Liu, Jing Shao and Dongrui Liu},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.2201},
  url = {https://doi.org/10.18653/v1/2026.acl-long.2201}
}

References

94 references, 46 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Cengiz Asmazoğlu Abdullah Bezir, Furkan Burhan Türkay. 2025a. Wiroai/wiroai-finance-llama-8b. find
  2. Cengiz Asmazoğlu Abdullah Bezir, Furkan Burhan Türkay. 2025b. Wiroai/wiroai-finance-qwen-7b. find
  3. Anton Alexandrov, Veselin Raychev, Mark Niklas Müller, Ce Zhang, Martin Vechev, and Kristina Toutanova. 2024. Mitigating catastrophic forgetting in language transfer via model merging. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 17167–17186, Miami, Florida, USA. A… find
  4. Yuntao Bai, Andy Jones, Kamal Ndousse, Amanda Askell, Anna Chen, Nova Dassarma, Dawn Drain, Stanislav Fort, Deep Ganguli, T. J. Henighan, Nicholas Joseph, Saurav Kadavath, John Kernion, Tom Conerly, Sheer El-Showk, Nelson Elhage, Zac Hatfield-Dodds, Danny Hernandez, Tristan Hume, and 12 others. 2022… find
  5. Shanghai AI Lab Yicheng Bao, Guanxu Chen, Mingkang Chen, Yunhao Chen, Chiyu Chen, Lingjie Chen, Sirui Chen, Xinquan Chen, Jie Cheng, Yu Cheng, Dengke Deng, Yizhuo Ding, Dan Ding, Xiaoshan Ding, Yizhuo Ding, Zhichen Dong, Lingxiao Du, Yu-Qi Fan, Xinshun Feng, and 97 others. 2025. Safework-r1: Coevolv… find
  6. Federico Bianchi, Mirac Suzgun, Giuseppe Attanasio, Paul Rottger, Dan Jurafsky, Tatsunori Hashimoto, and James Zou. 2024. Safety-tuned LLaMAs: Lessons from improving the safety of large language models that follow instructions. In The Twelfth International Conference on Learning Representations.
  7. Junying Chen, Zhenyang Cai, Ke Ji, Xidong Wang, Wanlong Liu, Rongsheng Wang, Jianye Hou, and Benyou Wang. 2024. Huatuogpt-o1, towards medical complex reasoning with llms. ArXiv, abs/2412.18925.
  8. Junying Chen, Zhenyang Cai, Ke Ji, Xidong Wang, Wanlong Liu, Rongsheng Wang, and Benyou Wang. 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… find
  9. Mark Chen, Jerry Tworek, Heewoo Jun, Qiming Yuan, Henrique Ponde De Oliveira Pinto, Jared Kaplan, Harri Edwards, Yuri Burda, Nicholas Joseph, Greg Brockman, and 1 others. 2021. Evaluating large language models trained on code. CoRR, abs/2107.03374.
  10. Aladin Djuhera, Swanand Kadhe, Farhan Ahmed, Syed Zawad, and Holger Boche. 2025. SafeMERGE: Preserving safety alignment in fine-tuned large language models via selective layer-wise model merging. In ICLR 2025 Workshop on Building Trust in Language Models and Applications. find
  11. Zeming Chen, Alejandro Hernández-Cano, Angelika Romanou, Antoine Bonnet, Kyle Matoba, Francesco Salvi, Matteo Pagliardini, Simin Fan, Andreas Köpf, Amirkeivan Mohtashami, Alexandre Sallinen, Alireza Sakhaeirad, Vinitra Swamy, Igor Krawczuk, Deniz Bayazit, Axel Marmet, Syrielle Montariol, Mary-Anne H… — MEDITRON-70B: Scaling Medical Pretraining for Large Language Models
  12. Clément Dumas, Chris Wendler, Veniamin Veselovsky, Giovanni Monea, and Robert West. 2025. Separating tongue from thought: Activation patching reveals language-agnostic concept representations in transformers. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics…
  13. Zhiyu Chen, Shiyang Li, Charese Smiley, Zhiqiang Ma, Sameena Shah, and William Yang Wang. 2022. Convfinqa: Exploring the chain of numerical reasoning in conversational finance question answering. Proceedings of EMNLP 2022.
  14. Bryan R Christ, Zachary Gottesman, Jonathan Kropko, and Thomas Hartvigsen. 2025. Math neurosurgery: Isolating language models’ math reasoning abilities using only forward passes. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages…
  15. Peter Clark, Isaac Cowhey, Oren Etzioni, Tushar Khot, Ashish Sabharwal, Carissa Schoenick, and Oyvind Tafjord. 2018. Think you have solved question answering? try arc, the AI2 reasoning challenge. CoRR, abs/1803.05457.
  16. 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. CoRR, abs/2110.14168.
  17. OpenCompass Contributors. 2023. Opencompass: A universal evaluation platform for foundation models. https://github.com/open-compass/ opencompass. ↗
  18. Damai Dai, Li Dong, Yaru Hao, Zhifang Sui, Baobao Chang, and Furu Wei. 2022. Knowledge neurons in pretrained transformers. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 8493– 8502, Dublin, Ireland. Association for Computatio… find
  19. Yunkai Dang, Kaichen Huang, Jiahao Huo, Yibo Yan, Sirui Huang, Dongrui Liu, Mengxi Gao, Jie Zhang, Chen Qian, Kun Wang, and 1 others. 2024. Explainable and interpretable multimodal large language Hugging Face. 2025. Open-r1: Fully open reproduction of deepseek-r1. Accessed: 2025-12-27. find
  20. Charles Goddard, Shamane Siriwardhana, Malikeh Ehghaghi, Luke Meyers, Vladimir Karpukhin, Brian Benedict, Mark McQuade, and Jacob Solawetz. 2024. Arcee’s MergeKit: A toolkit for merging large language models. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing:…
  21. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, and 1 others. 2024. The llama 3 herd of models. arXiv preprint arXiv:2407.21783.
  22. Maxime Griot, Coralie Hemptinne, Jean Vanderdonckt, and Demet Yuksel. 2025. Large language models lack essential metacognition for reliable medical reasoning. Nature Communications, 16(1). find
  23. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, and 1 others. 2025. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948.
  24. Shibo Hao, Sainbayar Sukhbaatar, DiJia Su, Xian Li, Zhiting Hu, Jason E Weston, and Yuandong Tian. 2025. Training large language model to reason in a continuous latent space. find
  25. Dan Hendrycks, Collin Burns, Steven Basart, Andrew Critch, Jerry Li, Dawn Song, and Jacob Steinhardt. 2021a. Aligning ai with shared human values. Proceedings of the International Conference on Learning Representations (ICLR).
  26. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2021b. Measuring massive multitask language understanding. Proceedings of the International Conference on Learning Representations (ICLR).
  27. Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. LoRA: Low-rank adaptation of large language models. In International Conference on Learning Representations.
  28. Xisen Jin, Xiang Ren, Daniel Preotiuc-Pietro, and Pengxiang Cheng. 2023. Dataless knowledge fusion by merging weights of language models. In The Eleventh International Conference on Learning Representations.
  29. Zhiyuan Hu, Yibo Wang, Hanze Dong, Yuhui Xu, Amrita Saha, Caiming Xiong, Bryan Hooi, and Junnan Li. 2025. Beyond’aha!’: Toward systematic meta-abilities alignment in large reasoning models. arXiv preprint arXiv:2505.10554. — [arXiv:2505.10554](https://arxiv.org/abs/2505.10554) find
  30. Chen Kai-tao, Ma Weijie, and Wang Xiaosong. 2025. Improving medical reasoning with curriculum-aware reinforcement learning. http://arxiv.org/abs/2505.19213. — [arXiv:2505.19213](https://arxiv.org/abs/2505.19213) ↗
  31. Gabriel Ilharco, Marco Tulio Ribeiro, Mitchell Wortsman, Ludwig Schmidt, Hannaneh Hajishirzi, and Ali Farhadi. 2023. Editing models with task arithmetic. In The Eleventh International Conference on Learning Representations.
  32. East Money Information. 2023. Openfindata. https://github.com/open-compass/OpenFinData/. ↗
  33. Pavel Izmailov, Dmitrii Podoprikhin, Timur Garipov, Dmitry P. Vetrov, and Andrew Gordon Wilson. 2018. Averaging weights leads to wider optima and better generalization. In Proceedings of the Thirty-Fourth Conference on Uncertainty in Artificial Intelligence, UAI 2018, Monterey, California, USA, Augu…
  34. Aaron Jaech, Adam Kalai, Adam Lerer, Adam Richardson, Ahmed El-Kishky, Aiden Low, Alec Helyar, Aleksander Madry, Alex Beutel, Alex Carney, and 1 others. 2024. Openai o1 system card. arXiv preprint arXiv:2412.16720.
  35. Naman Jain, King Han, Alex Gu, Wen-Ding Li, Fanjia Yan, Tianjun Zhang, Sida I. Wang, Armando Solar-Lezama, Koushik Sen, and Ion Stoica. 2024. Livecodebench: Holistic and contamination free evaluation of large language models for code. arXiv preprint arXiv:2403.07974.
  36. Fengqing Jiang, Zhangchen Xu, Yuetai Li, Luyao Niu, Zhen Xiang, Bo Li, Bill Yuchen Lin, and Radha Poovendran. 2025. Safechain: Safety of language models with long chain-of-thought reasoning capabilities. arXiv preprint arXiv:2502.12025. find
  37. Di Jin, Eileen Pan, Nassim Oufattole, Wei-Hung Weng, Hanyi Fang, and Peter Szolovits. 2020. What disease does this patient have? A large-scale open domain question answering dataset from medical exams. CoRR, abs/2009.13081.
  38. Qiao Jin, Bhuwan Dhingra, Zhengping Liu, William W. Cohen, and Xinghua Lu. 2019. Pubmedqa: A dataset for biomedical research question answering. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Pr…
  39. Martin Kuo, Jianyi Zhang, Aolin Ding, Qinsi Wang, Louis DiValentin, Yujia Bao, Wei Wei, Hai Li, and Yiran Chen. 2025. H-cot: Hijacking the chain-of-thought safety reasoning mechanism to jailbreak large reasoning models, including openai o1/o3, deepseek-r1, and gemini 2.0 flash thinking. arXiv prepri… — [arXiv:2502.12893](https://arxiv.org/abs/2502.12893) find
  40. Xiaochong Lan, Yu Zheng, Shiteng Cao, and Yong Li. 2025. The thinking spectrum: An empirical study of tunable reasoning in llms through model merging. arXiv preprint arXiv:2509.22034. — [arXiv:2509.22034](https://arxiv.org/abs/2509.22034) find
  41. Ming Li, Yanhong Li, and Tianyi Zhou. 2025a. What happened in LLMs layers when trained for fast vs. slow thinking: A gradient perspective. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 32017–32154, Vienna, Austria. Associati…
  42. Weishi Li, Yong Peng, Miao Zhang, Liang Ding, Han Hu, and Li Shen. 2023. Deep model fusion: A survey. CoRR, abs/2309.15698. find
  43. Yunshui Li, Yiyuan Ma, Shen Yan, Chaoyi Zhang, Jing Liu, Jianqiao Lu, Ziwen Xu, Mengzhao Chen, Minrui Wang, Shiyi Zhan, Jin Ma, Xunhao Lai, Yao Luo, Xingyan Bin, Hongbin Ren, Mingji Han, Wenhao Hao, Bairen Yi, LingJun Liu, and 5 others. 2025b. Model merging in pre-training of large language models.… find
  44. 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. arXiv preprint arXiv:2305.20050.
  45. Shuqi Liu, Han Wu, Bowei He, Xiongwei Han, Mingxuan Yuan, and Linqi Song. 2025. Sens-merging: Sensitivity-guided parameter balancing for merging large language models. CoRR, abs/2502.12420. find
  46. Jinliang Lu, Ziliang Pang, Min Xiao, Yaochen Zhu, Rui Xia, and Jiajun Zhang. 2024. Merge, ensemble, and cooperate! A survey on collaborative strategies in the era of large language models. CoRR, abs/2407.06089. — [arXiv:2407.06089](https://arxiv.org/abs/2407.06089) find
  47. Qianli Ma, Dongrui Liu, Chen Qian, Linfeng Zhang, and Jing Shao. 2025. Led-merging: Mitigating safety-utility conflicts in model merging with location-election-disjoint. CoRR.
  48. Spencer Mateega, Carlos Georgescu, and Danny Tang. 2025. Financeqa: A benchmark for evaluating financial analysis capabilities of large language models. ArXiv, abs/2501.18062. — [arXiv:2501.18062](https://arxiv.org/abs/2501.18062) find
  49. Michael Matena and Colin Raffel. 2022. Merging models with fisher-weighted averaging. In Advances in Neural Information Processing Systems 35: Annual Conference on Neural Information Processing Systems 2022, NeurIPS 2022, New Orleans, LA, USA, November 28 - December 9, 2022.
  50. Kohsei Matsutani, Shota Takashiro, Kojima Takeshi, Gouki Minegishi, Yusuke Iwasawa, Takeshi Kojima, Yutaka Matsuo, Yusuke Iwasawa, and Yutaka Matsuo. 2025. Rl squeezes, sft expands: A comparative study of reasoning llms. http://arxiv.org/abs/2509.21128, abs/2509.21128. — [arXiv:2509.21128](https://arxiv.org/abs/2509.21128) ↗
  51. Mantas Mazeika, Long Phan, Xuwang Yin, Andy Zou, Zifan Wang, Norman Mu, Elham Sakhaee, Nathaniel Li, Steven Basart, Bo Li, David Forsyth, and Dan Hendrycks. 2024. Harmbench: A standardized evaluation framework for automated red teaming and robust refusal.
  52. OpenAI. 2025. Openai o3 and o4-mini system card. find
  53. German I. Parisi, Ronald Kemker, Jose L. Part, Christopher Kanan, and Stefan Wermter. 2019. Continual lifelong learning with neural networks: A review. Neural Networks, 113:54–71.
  54. Venkatesh Balavadhani Parthasarathy, Ahtsham Zafar, Aafaq Khan, and Arsalan Shahid. 2024. The ultimate guide to fine-tuning llms from basics to breakthroughs: An exhaustive review of technologies, research, best practices, applied research challenges and opportunities. arXiv preprint arXiv:2408.1329… — [arXiv:2408.13296](https://arxiv.org/abs/2408.13296) find
  55. Kunat Pipatanakul, Pittawat Taveekitworachai, Potsawee Manakul, and Kasima Tharnpipitchai. 2025. Adapting language-specific llms to a reasoning model in one day via model merging–an open recipe. arXiv preprint arXiv:2502.09056. — [arXiv:2502.09056](https://arxiv.org/abs/2502.09056) find
  56. Saif Punjwani and Larry Heck. 2025. Weight-of-thought reasoning: Exploring neural network weights for enhanced llm reasoning. arXiv preprint arXiv:2504.10646. find
  57. Chen Qian, Dongrui Liu, Jie Zhang, Yong Liu, and Jing Shao. 2025a. The tug of war within: Mitigating the fairness-privacy conflicts in large language models. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 12066–12095. Associa… find
  58. Lingfei Qian, Weipeng Zhou, Yan Wang, Xueqing Peng, Jimin Huang, and Qianqian Xie. 2025b. Fino1: On the transferability of reasoning enhanced llms to finance. Preprint, arXiv:2502.08127. — [arXiv:2502.08127](https://arxiv.org/abs/2502.08127) find
  59. Pengcheng Qiu, Chaoyi Wu, Xiaoman Zhang, Weixiong Lin, Haicheng Wang, Ya Zhang, Yanfeng Wang, and Weidi Xie. 2024. Towards building multilingual language model for medicine. Preprint, arXiv:2402.13963.
  60. Daking Rai and Ziyu Yao. 2024. An investigation of neuron activation as a unified lens to explain chain-of-thought eliciting arithmetic reasoning of LLMs. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 7174–7193, Bangkok, Tha… find
  61. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R. Bowman. 2023. GPQA: A graduate-level google-proof q&a benchmark. arXiv preprint arXiv:2311.12022.
  62. Wei Ruan, Tianze Yang, Yifan Zhou, Tianming Liu, and Jin Lu. 2025. From task-specific models to unified systems: A review of model merging approaches. CoRR, abs/2503.08998. — [arXiv:2503.08998](https://arxiv.org/abs/2503.08998) find
  63. Alessandro Stolfo, Yonatan Belinkov, and Mrinmaya Sachan. 2023. A mechanistic interpretation of arithmetic reasoning in language models using causal mediation analysis. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 7035–7052, Singapore. Association… find
  64. Tianyi Tang, Wenyang Luo, Haoyang Huang, Dongdong Zhang, Xiaolei Wang, Xin Zhao, Furu Wei, and Ji-Rong Wen. 2024. Language-specific neurons: The key to multilingual capabilities in large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volu…
  65. Xinyu Tang, Xiaolei Wang, Zhihao Lv, Yingqian Min, Xin Zhao, Binbin Hu, Ziqi Liu, and Zhiqiang Zhang. 2025a. Unlocking general long chain-of-thought reasoning capabilities of large language models via representation engineering. In Proceedings of the 63rd Annual Meeting of the Association for Comput…
  66. Yiru Tang, Kun Zhou, Yingqian Min, Xin Zhao, Jing Sha, Zhichao Sheng, and Shijin Wang. 2025b. Enhancing chain-of-thought reasoning via neuron activation differential analysis. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 16162–16170, Suzhou, China.… find
  67. P1 Team. 2025a. P1: Mastering physics olympiads with reinforcement learning. find
  68. Qwen Team. 2025b. Qwq-32b: Embracing the power of reinforcement learning. find
  69. Rahul Thapa, Qingyang Wu, Kevin Wu, Harrison Zhang, Angela Zhang, Eric Wu, Haotian Ye, Suhana Bedi, Nevin Aresh, Joseph Boen, Shriya Reddy, Ben Athiwaratkun, Shuaiwen Leon Song, and James Zou. 2025. Disentangling reasoning and knowledge in medical large language models. ArXiv, abs/2505.11462. — [arXiv:2505.11462](https://arxiv.org/abs/2505.11462) find
  70. Ehsan Ullah, Anil Parwani, Mirza Baig, and Rajendra Singh. 2024. Challenges and barriers of using large language models (llm) such as chatgpt for diagnostic medicine with a focus on digital pathology – a recent scoping review. Diagnostic Pathology, 19(1). find
  71. Hemish Veeraboina. 2023. Aime problem set 1983-2024. find
  72. Fanqi Wan, Xinting Huang, Deng Cai, Xiaojun Quan, Wei Bi, and Shuming Shi. 2024. Knowledge fusion of large language models. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024. Open-Review.net.
  73. Eric J. Wang. 2023. Alpaca-lora. https://github.com/tloen/alpaca-lora. ↗
  74. Xiaozhi Wang, Kaiyue Wen, Zhengyan Zhang, Lei Hou, Zhiyuan Liu, and Juanzi Li. 2022. Finding skill neurons in pre-trained transformer-based language models. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 11132–11152, Abu Dhabi, United Arab Emirates.…
  75. Somshubra Majumdar Aleksander Ficek Siddhartha Jain Jocelyn Huang Vahid Noroozi Boris Ginsburg Wasi Uddin Ahmad, Sean Narenthiran. 2025. Opencodereasoning: Advancing data distillation for competitive coding.
  76. Xuansheng Wu, Haiyan Zhao, Yaochen Zhu, Yucheng Shi, Fan Yang, Lijie Hu, Tianming Liu, Xiaoming Zhai, Wenlin Yao, Jundong Li, and 1 others. 2024. Usable xai: 10 strategies towards exploiting explainability in the llm era. arXiv preprint arXiv:2403.08946. — [arXiv:2403.08946](https://arxiv.org/abs/2403.08946) find
  77. Prateek Yadav, Derek Tam, Leshem Choshen, Colin Raffel, and Mohit Bansal. 2023. Ties-merging: Resolving interference when merging models. In Thirty-seventh Conference on Neural Information Processing Systems.
  78. 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, Junyang Lin, Kai Dang, and 23 others. 2024a. Qwen2.5 technical report. arXiv preprint arX… find
  79. Enneng Yang, Li Shen, Guibing Guo, Xingwei Wang, Xiaochun Cao, Jie Zhang, and Dacheng Tao. 2024b. Model merging in llms, mllms, and beyond: Methods, theories, applications and opportunities. arXiv preprint arXiv:2408.07666.
  80. Enneng Yang, Li Shen, Guibing Guo, Xingwei Wang, Xiaochun Cao, Jie Zhang, and Dacheng Tao. 2024c. Model merging in llms, mllms, and beyond: Methods, theories, applications and opportunities. arXiv preprint arXiv:2408.07666.
  81. Enneng Yang, Zhenyi Wang, Li Shen, Shiwei Liu, Guibing Guo, Xingwei Wang, and Dacheng Tao. 2024d. Adamerging: Adaptive model merging for multi-task learning. In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024. OpenReview.net.
  82. Jinluan Yang, Anke Tang, Didi Zhu, Zhengyu Chen, Li Shen, and Fei Wu. 2025a. Mitigating the backdoor effect for multi-task model merging via safety-aware subspace. In The Thirteenth International Conference on Learning Representations. find
  83. Junyao Yang, Jianwei Wang, Huiping Zhuang, Cen Chen, and Ziqian Zeng. 2025b. Rcp-merging: Merging long chain-of-thought models with domain-specific models by considering reasoning capability as prior. arXiv preprint arXiv:2508.03140. — [arXiv:2508.03140](https://arxiv.org/abs/2508.03140) find
  84. Edward Yeo, Yuxuan Tong, Morry Niu, Graham Neubig, and Xiang Yue. 2025. Demystifying long chain-of-thought reasoning in llms. Preprint, arXiv:2502.03373.
  85. Le Yu, Bowen Yu, Haiyang Yu, Fei Huang, and Yongbin Li. 2024. Language models are super mario: Absorbing abilities from homologous models as a free lunch. Preprint, arXiv:2311.03099.
  86. Zeping Yu, Yonatan Belinkov, and Sophia Ananiadou. 2025. Back attention: Understanding and enhancing multi-hop reasoning in large language models. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 11268–11283, Suzhou, China. Association for Computationa…
  87. Nie Yuqi, Yaxuan Kong, Xiaowen Dong, John Mulvey, Vincent Poor, Qingsong Wen, and Stefan Zohren. 2024. A survey of large language models for financial applications: Progress, prospects and challenges. arXiv (Cornell University). find
  88. Mohammad Zbeeb, Hasan Abed Al Kader Hammoud, and Bernard Ghanem. 2025. Reasoning vectors: Transferring chain-of-thought capabilities via task arithmetic. arXiv preprint arXiv:2509.01363. — [arXiv:2509.01363](https://arxiv.org/abs/2509.01363) find
  89. Zhiyuan Zeng, Qinyuan Cheng, Zhangyue Yin, Yunhua Zhou, and Xipeng Qiu. 2025. Revisiting the test-time scaling of o1-like models: Do they truly possess test-time scaling capabilities? CoRR, abs/2502.12215.
  90. Chongwen Zhao and Kaizhu Huang. 2025. Unraveling llm jailbreaks through safety knowledge neurons. arXiv preprint arXiv:2509.01631.
  91. Huaqin Zhao, Zhengliang Liu, Zihao Wu, Yiwei Li, Tianze Yang, Peng Shu, Shaochen Xu, Haixing Dai, Lin Zhao, Gengchen Mai, Ninghao Liu, and Liu Tianming. 2024. Revolutionizing finance with llms: An overview of applications and insights. arXiv (Cornell University).
  92. Qi Zhou, Yiming Zhang, Yanggan Gu, Yuanyi Wang, Zhijie Sang, Zhaoyi Yan, Zhen Li, Shengyu Zhang, Fei Wu, and Hongxia Yang. 2025. Democratizing ai through model fusion: A comprehensive review and future directions. Nexus, 2(4):100102. find
  93. Andy Zou, Zifan Wang, J. Zico Kolter, and Matt Fredrikson. 2023. Universal and transferable adversarial attacks on aligned language models. Preprint, arXiv:2307.15043.
  94. Theoretical Justification for Gradient Magnitude Metrics find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/yang-2026-reasonany-incorporating-reasoning-capability-any/