---
title: "Decoupling Generalization and Adaptation in Meta-Learning for"
description: "The authors propose DeGAML-LLM, which separates the two roles. 41 references. Among them: 0."
authors: ["Nitin Vetcha", "Binqian Xu", "Dianbo Liu"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)"
source: "https://aclanthology.org/2026.acl-short.69/"
doi: "10.18653/v1/2026.acl-short.69"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/vetcha-2026-decoupling-generalization-adaptation-meta-learning/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/vetcha-2026-decoupling-generalization-adaptation-meta-learning/)

[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/27/) / Decoupling Generalization and Adaptation in Meta-Learning for Large Language Models

# Decoupling Generalization and Adaptation in Meta-Learning for Large Language Models

**Authors:** Nitin Vetcha, Binqian Xu and Dianbo Liu · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)](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-short.69/) · [PDF](https://aclanthology.org/2026.acl-short.69.pdf) · [DOI 10.18653/v1/2026.acl-short.69](https://doi.org/10.18653/v1/2026.acl-short.69) · [OpenAlex](https://openalex.org/W7166837157)

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

## Summary
Meta-learning methods for LLMs such as MAML-en-LLM and ABMLL help LoRA-adapted models learn new tasks quickly. However, they optimize a shared initialization and per-task adjustment jointly, even though these goals conflict, and they allow only gradient-based adaptation. The authors propose DeGAML-LLM, which separates the two roles. A generalization module, a task-conditioned generator, reads a small support set and outputs a distribution over LoRA adapter weights. An independent adaptation policy, trained with reinforcement learning (ReST^EM), then perturbs those weights step by step and is rewarded with a binary signal when validation performance improves. This means the adaptation stage does not need gradients from the meta-objective. Experiments use Qwen2.5-1.5B-Instruct and Qwen2.5-0.5B-Instruct on six in-domain common-sense and QA benchmarks and six out-of-domain ones, with only 20 target examples for adaptation. DeGAML-LLM achieves the best average score in all four model/domain combinations, though not on every individual benchmark. The results suggest that decoupling the two roles makes few-shot LLM adaptation more flexible, including with non-differentiable feedback.

## Key points
- Baselines are ABMLL, MAML-en-LLM, "No Meta-Train LoRA" and "Union Train LoRA". The adaptation setting follows ABMLL and uses 20 datapoints from the target dataset.
- The in-domain tasks are ARC-c, ARC-e, HellaSwag, BoolQ, PIQA and WinoGrande. The out-of-domain tasks are GSM-8K, MATH, DivLogicEval, SocialIQA, CodeMMLU and JAMA Clinical.
- In-domain average:
  - Qwen2.5-1.5B-Instruct scores 67.7, versus 65.5 for MAML-en-LLM and 62.4 for ABMLL.
  - Qwen2.5-0.5B-Instruct scores 58.4, versus 51.9 and 42.1.
- Out-of-domain average:
  - The 1.5B model scores 47.5, versus 42.3 for MAML-en-LLM and 34.3 for ABMLL.
  - The 0.5B model scores 34.2, versus 32.6 and 24.9.
  - The method loses on some tasks, for example BoolQ and MATH with the 0.5B model.
- The ablation shows that RL adaptation contributes most of the gain. For the 1.5B model, the in-domain average rises from 59.6 (base) to 62.5 with the generator alone and to 67.7 with adaptation. For the 0.5B model, it rises from 37.2 to 45.7 and then 58.4.

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

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works

🧭 Topics (1): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works

## Cite

Cite as: Nitin Vetcha, Binqian Xu and Dianbo Liu. 2026. Decoupling Generalization and Adaptation in Meta-Learning for Large Language Models. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers).

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

```bibtex
@inproceedings{vetcha2026decoupling,
  title = {{Decoupling Generalization and Adaptation in Meta-Learning for Large Language Models}},
  author = {Nitin Vetcha, Binqian Xu and Dianbo Liu},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-short.69},
  url = {https://doi.org/10.18653/v1/2026.acl-short.69}
}
```

- [← previous Deconstructing Instruction-Following: A New Benchmark for Granular Evaluation of Large Language Model Instruction Compliance Abilities](https://md.research.elamur.ai/free/en/papers/purpura-2026-deconstructing-instruction-following-new-benchmark/)
- [next → DeepICD-R1: Medical Reasoning through Hierarchical Rewards and Unsupervised Distillation](https://md.research.elamur.ai/free/en/papers/rohr-2026-deepicd-r1-medical-reasoning-through/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/27/)

## References

41 references, 15 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Aaron Grattafiori and 1 others. 2024. [The llama 3 herd of models](https://md.research.elamur.ai/free/en/papers/grattafiori-2024-llama-3-herd-models/). arXiv preprint arXiv:2407.21783.
2. Josh Achiam, Steven Adler, Sandhini Agarwal, and 1 others. 2023. [Gpt-4 technical report](https://md.research.elamur.ai/free/en/papers/openai-2023-gpt-4-technical-report/). arXiv preprint arXiv:2303.08774.
3. Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. 2021. [Measuring mathematical problem solving with the math dataset](https://md.research.elamur.ai/free/en/papers/hendrycks-2021-measuring-mathematical-problem-solving-math/). Preprint, arXiv:2103.03874.
4. Yonatan Bisk, Rowan Zellers, Ronan Le Bras, Jianfeng Gao, and Yejin Choi. 2019. [Piqa: Reasoning about physical commonsense in natural language](https://md.research.elamur.ai/free/en/papers/bisk-2019-piqa-reasoning-about-physical-commonsense/). Preprint, arXiv:1911.11641.
5. Hanjie Chen, Zhouxiang Fang, Yash Singla, and Mark Dredze. 2025. [Benchmarking large language models on answering and explaining challenging medical questions](https://md.research.elamur.ai/free/en/papers/chen-2025-benchmarking-large-language-models-answering/). In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language…
6. Yanda Chen, Ruiqi Zhong, Sheng Zha, George Karypis, and He He. 2022. Meta-learning via language model in-context tuning. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Yanda+Chen%2C+Ruiqi+Zhong%2C+Sheng+Zha%2C+George+Karypis%2C+and+He+He%2E+2022%2E+Meta%2Dlearning+via+language+model+in%2Dcontext+tuning%2E+In+Proceedings+of+the+60th+Annual+Meeting+of+the+Association+for+Computational+)
7. Tsz Ting Chung, Lemao Liu, Mo Yu, and Dit-Yan Yeung. 2025. Divlogiceval: A framework for benchmarking logical reasoning evaluation in large language models. Preprint, arXiv:2509.15587. [find](https://scholar.google.com/scholar?q=Tsz+Ting+Chung%2C+Lemao+Liu%2C+Mo+Yu%2C+and+Dit%2DYan+Yeung%2E+2025%2E+Divlogiceval%3A+A+framework+for+benchmarking+logical+reasoning+evaluation+in+large+language+models%2E+Preprint%2C+arXiv%3A2509%2E15587%2E)
8. Timothy Hospedales, Antreas Antoniou, Paul Micaelli, and Amos Storkey. 2021. Meta-learning in neural networks: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(9):5149– 5169. [find](https://scholar.google.com/scholar?q=Timothy+Hospedales%2C+Antreas+Antoniou%2C+Paul+Micaelli%2C+and+Amos+Storkey%2E+2021%2E+Meta%2Dlearning+in+neural+networks%3A+A+survey%2E+IEEE+Transactions+on+Pattern+Analysis+and+Machine+Intelligence%2C+44%289%29%3A5149%E2%80%93+516)
9. Zejiang Hou, Julian Salazar, and George Polovets. 2022. Meta-learning the difference: preparing large language models for efficient adaptation. Transactions of the Association for Computational Linguistics, 10:1249–1265. [find](https://scholar.google.com/scholar?q=Zejiang+Hou%2C+Julian+Salazar%2C+and+George+Polovets%2E+2022%2E+Meta%2Dlearning+the+difference%3A+preparing+large+language+models+for+efficient+adaptation%2E+Transactions+of+the+Association+for+Computational+Lingui)
10. Neil Houlsby, Andrei Giurgiu, Stanislaw Jastrzebski, and 1 others. 2019. Parameter-efficient transfer learning for nlp. In International Conference on Machine Learning. [find](https://scholar.google.com/scholar?q=Neil+Houlsby%2C+Andrei+Giurgiu%2C+Stanislaw+Jastrzebski%2C+and+1+others%2E+2019%2E+Parameter%2Defficient+transfer+learning+for+nlp%2E+In+International+Conference+on+Machine+Learning%2E)
11. Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2021. [Lora: Low-rank adaptation of large language models](https://md.research.elamur.ai/free/en/papers/hu-2021-lora-low-rank-adaptation-large/). arXiv preprint arXiv:2106.09685.
12. Christopher Clark, Kenton Lee, Ming-Wei Chang, Tom Kwiatkowski, Michael Collins, and Kristina Toutanova. 2019. [BoolQ: Exploring the surprising difficulty of natural yes/no questions](https://md.research.elamur.ai/free/en/papers/clark-2019-boolq-exploring-surprising-difficulty-natural/). In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics:…
13. Jinwu Hu, Zitian Zhang, Guohao Chen, Xutao Wen, Chao Shuai, Wei Luo, Bin Xiao, Yuanqing Li, and Mingkui Tan. 2025a. Test-time learning for large language models. In Forty-second International Conference on Machine Learning. [find](https://scholar.google.com/scholar?q=Jinwu+Hu%2C+Zitian+Zhang%2C+Guohao+Chen%2C+Xutao+Wen%2C+Chao+Shuai%2C+Wei+Luo%2C+Bin+Xiao%2C+Yuanqing+Li%2C+and+Mingkui+Tan%2E+2025a%2E+Test%2Dtime+learning+for+large+language+models%2E+In+Forty%2Dsecond+International+Conferen)
14. 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](https://md.research.elamur.ai/free/en/papers/clark-2018-think-have-solved-question-answering/). Preprint, arXiv:1803.05457.
15. Minjun Kim, Jeongwon Park, Seonghun Kang, and Bongkyoung Kwon. 2025. Reptilora: accelerating meta-initialization with low-rank adaptation. Issues in Information Systems, 26(3). [find](https://scholar.google.com/scholar?q=Minjun+Kim%2C+Jeongwon+Park%2C+Seonghun+Kang%2C+and+Bongkyoung+Kwon%2E+2025%2E+Reptilora%3A+accelerating+meta%2Dinitialization+with+low%2Drank+adaptation%2E+Issues+in+Information+Systems%2C+26%283%29%2E)
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](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/). Preprint, arXiv:2110.14168.
17. Julian Coda-Forno, Marcel Binz, Zeynep Akata, Matt Botvinick, Jane Wang, and Eric Schulz. 2023. Meta-in-context learning in large language models. Advances in Neural Information Processing Systems, 36:65189–65201. [find](https://scholar.google.com/scholar?q=Julian+Coda%2DForno%2C+Marcel+Binz%2C+Zeynep+Akata%2C+Matt+Botvinick%2C+Jane+Wang%2C+and+Eric+Schulz%2E+2023%2E+Meta%2Din%2Dcontext+learning+in+large+language+models%2E+Advances+in+Neural+Information+Processing+Systems%2C+36)
18. Chelsea Finn, Pieter Abbeel, and Sergey Levine. 2017. [Model-agnostic meta-learning for fast adaptation of deep networks](https://md.research.elamur.ai/free/en/papers/finn-2017-model-agnostic-meta-learning-fast/). In International Conference on Machine Learning, pages 1126–1135.
19. Yang Hu, Xingyu Zhang, Xueji Fang, Zhiyang Chen, Xiao Wang, Huatian Zhang, and Guojun Qi. 2025b. Slot: Sample-specific language model optimization at test-time. arXiv preprint arXiv:2505.12392. — [arXiv:2505.12392](https://arxiv.org/abs/2505.12392) [find](https://scholar.google.com/scholar?q=Yang+Hu%2C+Xingyu+Zhang%2C+Xueji+Fang%2C+Zhiyang+Chen%2C+Xiao+Wang%2C+Huatian+Zhang%2C+and+Guojun+Qi%2E+2025b%2E+Slot%3A+Sample%2Dspecific+language+model+optimization+at+test%2Dtime%2E+arXiv+preprint+arXiv%3A2505%2E12392%2E+%E2%80%94+%5BarX)
20. Minyoung Kim and Timothy Hospedales. 2025. Lift: Learning to fine-tune via bayesian parameter efficient meta fine-tuning. In The Thirteenth International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Minyoung+Kim+and+Timothy+Hospedales%2E+2025%2E+Lift%3A+Learning+to+fine%2Dtune+via+bayesian+parameter+efficient+meta+fine%2Dtuning%2E+In+The+Thirteenth+International+Conference+on+Learning+Representations%2E)
21. Suhas Kotha, Jacob Mitchell Springer, and Aditi Raghunathan. 2024. [Understanding catastrophic forgetting in language models via implicit inference](https://md.research.elamur.ai/free/en/papers/kotha-2023-understanding-catastrophic-forgetting-language-models/). In ICLR.
22. Zhiyuan Liang, Dongwen Tang, Yuhao Zhou, Xuanlei Zhao, Mingjia Shi, Wangbo Zhao, Zekai Li, Peihao Wang, Konstantin Schürholt, Damian Borth, and 1 others. 2025. Drag-and-drop llms: Zero-shot prompt-to-weights. arXiv preprint arXiv:2506.16406. — [arXiv:2506.16406](https://arxiv.org/abs/2506.16406) [find](https://scholar.google.com/scholar?q=Zhiyuan+Liang%2C+Dongwen+Tang%2C+Yuhao+Zhou%2C+Xuanlei+Zhao%2C+Mingjia+Shi%2C+Wangbo+Zhao%2C+Zekai+Li%2C+Peihao+Wang%2C+Konstantin+Sch%C3%BCrholt%2C+Damian+Borth%2C+and+1+others%2E+2025%2E+Drag%2Dand%2Ddrop+llms%3A+Zero%2Dshot+prompt%2Dto%2D)
23. Yun Luo, Zhen Yang, Fandong Meng, Yafu Li, Jie Zhou, and Yue Zhang. 2025. An empirical study of catastrophic forgetting in large language models during continual fine-tuning. IEEE Transactions on Audio, Speech and Language Processing. [find](https://scholar.google.com/scholar?q=Yun+Luo%2C+Zhen+Yang%2C+Fandong+Meng%2C+Yafu+Li%2C+Jie+Zhou%2C+and+Yue+Zhang%2E+2025%2E+An+empirical+study+of+catastrophic+forgetting+in+large+language+models+during+continual+fine%2Dtuning%2E+IEEE+Transactions+on+Audi)
24. Dung Nguyen Manh, Thang Phan Chau, Nam Le Hai, Thong T. Doan, Nam V. Nguyen, Quang Pham, and Nghi D. Q. Bui. 2025. Codemmlu: A multi-task benchmark for assessing code understanding & reasoning capabilities of codellms. Preprint, arXiv:2410.01999. — [arXiv:2410.01999](https://arxiv.org/abs/2410.01999) [find](https://scholar.google.com/scholar?q=Dung+Nguyen+Manh%2C+Thang+Phan+Chau%2C+Nam+Le+Hai%2C+Thong+T%2E+Doan%2C+Nam+V%2E+Nguyen%2C+Quang+Pham%2C+and+Nghi+D%2E+Q%2E+Bui%2E+2025%2E+Codemmlu%3A+A+multi%2Dtask+benchmark+for+assessing+code+understanding+%26+reasoning+capabil)
25. Sewon Min, Mike Lewis, Luke Zettlemoyer, and Hannaneh Hajishirzi. 2022. Metaicl: Learning to learn in context. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Sewon+Min%2C+Mike+Lewis%2C+Luke+Zettlemoyer%2C+and+Hannaneh+Hajishirzi%2E+2022%2E+Metaicl%3A+Learning+to+learn+in+context%2E+In+Proceedings+of+the+2022+Conference+of+the+North+American+Chapter+of+the+Association+fo)
26. Alex Nichol, Joshua Achiam, and John Schulman. 2018. On first-order meta-learning algorithms. arXiv preprint arXiv:1803.02999. [find](https://scholar.google.com/scholar?q=Alex+Nichol%2C+Joshua+Achiam%2C+and+John+Schulman%2E+2018%2E+On+first%2Dorder+meta%2Dlearning+algorithms%2E+arXiv+preprint+arXiv%3A1803%2E02999%2E)
27. Keisuke Sakaguchi, Ronan Le Bras, Chandra Bhagavatula, and Yejin Choi. 2021. Winogrande: an adversarial winograd schema challenge at scale. Commun. ACM, 64(9):99–106. [find](https://scholar.google.com/scholar?q=Keisuke+Sakaguchi%2C+Ronan+Le+Bras%2C+Chandra+Bhagavatula%2C+and+Yejin+Choi%2E+2021%2E+Winogrande%3A+an+adversarial+winograd+schema+challenge+at+scale%2E+Commun%2E+ACM%2C+64%289%29%3A99%E2%80%93106%2E)
28. Maarten Sap, Hannah Rashkin, Derek Chen, Ronan LeBras, and Yejin Choi. 2019. [Socialiqa: Commonsense reasoning about social interactions](https://md.research.elamur.ai/free/en/papers/sap-2019-socialiqa-commonsense-reasoning-about-social/). Preprint, arXiv:1904.09728.
29. Jürgen Schmidhuber. 1987. Evolutionary principles in self-referential learning. Diploma thesis, Technische Universität München. [find](https://scholar.google.com/scholar?q=J%C3%BCrgen+Schmidhuber%2E+1987%2E+Evolutionary+principles+in+self%2Dreferential+learning%2E+Diploma+thesis%2C+Technische+Universit%C3%A4t+M%C3%BCnchen%2E)
30. Avi Singh, John D. Co-Reyes, Rishabh Agarwal, Ankesh Anand, Piyush Patil, Xavier Garcia, Peter J. Liu, James Harrison, Jaehoon Lee, Kelvin Xu, Aaron Parisi, Abhishek Kumar, Alex Alemi, Alex Rizkowsky, Azade Nova, Ben Adlam, Bernd Bohnet, Gamaleldin Elsayed, Hanie Sedghi, and 22 others. 2024. [Beyond…](https://md.research.elamur.ai/free/en/papers/singh-2023-human-data-scaling-self-training/)
31. Sanchit Sinha, Yuguang Yue, Victor Soto, Mayank Kulkarni, Jianhua Lu, and Aidong Zhang. 2024. Maml-en-llm: Model agnostic meta-training of llms for improved in-context learning. In ACM SIGKDD Conference on Knowledge Discovery and Data Mining. [find](https://scholar.google.com/scholar?q=Sanchit+Sinha%2C+Yuguang+Yue%2C+Victor+Soto%2C+Mayank+Kulkarni%2C+Jianhua+Lu%2C+and+Aidong+Zhang%2E+2024%2E+Maml%2Den%2Dllm%3A+Model+agnostic+meta%2Dtraining+of+llms+for+improved+in%2Dcontext+learning%2E+In+ACM+SIGKDD+Conferen)
32. Jake Snell, Kevin Swersky, and Richard Zemel. 2017. Prototypical networks for few-shot learning. In Advances in Neural Information Processing Systems. [find](https://scholar.google.com/scholar?q=Jake+Snell%2C+Kevin+Swersky%2C+and+Richard+Zemel%2E+2017%2E+Prototypical+networks+for+few%2Dshot+learning%2E+In+Advances+in+Neural+Information+Processing+Systems%2E)
33. Sebastian Thrun and Lorien Pratt. 1998. Learning to learn: Introduction and overview. Learning to Learn, pages 3–17. [find](https://scholar.google.com/scholar?q=Sebastian+Thrun+and+Lorien+Pratt%2E+1998%2E+Learning+to+learn%3A+Introduction+and+overview%2E+Learning+to+Learn%2C+pages+3%E2%80%9317%2E)
34. Ricardo Vilalta and Youssef Drissi. 2002. A perspective view and survey of meta-learning. Artificial Intelligence Review, 18(2):77–95. [find](https://scholar.google.com/scholar?q=Ricardo+Vilalta+and+Youssef+Drissi%2E+2002%2E+A+perspective+view+and+survey+of+meta%2Dlearning%2E+Artificial+Intelligence+Review%2C+18%282%29%3A77%E2%80%9395%2E)
35. Oriol Vinyals, Charles Blundell, Timothy Lillicrap, Daan Wierstra, and 1 others. 2016. [Matching networks for one shot learning](https://md.research.elamur.ai/free/en/papers/vinyals-2016-matching-networks-one-shot-learning/). In Advances in Neural Information Processing Systems, pages 3630–3638.
36. Kai Wang and 1 others. 2025. Recurrent parameter generation. arXiv preprint. [find](https://scholar.google.com/scholar?q=Kai+Wang+and+1+others%2E+2025%2E+Recurrent+parameter+generation%2E+arXiv+preprint%2E)
37. Taiqiang Wu, Jiahao Wang, Zhe Zhao, and Ngai Wong. 2024. Mixture-of-subspaces in low-rank adaptation. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 7880–7899. [find](https://scholar.google.com/scholar?q=Taiqiang+Wu%2C+Jiahao+Wang%2C+Zhe+Zhao%2C+and+Ngai+Wong%2E+2024%2E+Mixture%2Dof%2Dsubspaces+in+low%2Drank+adaptation%2E+In+Proceedings+of+the+2024+Conference+on+Empirical+Methods+in+Natural+Language+Processing%2C+pages+7)
38. An Yang, Baosong Yang, Binyuan Hui, and 1 others. 2024. [Qwen2.5 technical report](https://md.research.elamur.ai/free/en/papers/hui-2024-qwen2-5-coder-technical-report/). arXiv preprint arXiv:2409.12186.
39. Rowan Zellers, Ari Holtzman, Yonatan Bisk, Ali Farhadi, and Yejin Choi. 2019. HellaSwag: Can a machine really finish your sentence? In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 4791–4800, Florence, Italy. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Rowan+Zellers%2C+Ari+Holtzman%2C+Yonatan+Bisk%2C+Ali+Farhadi%2C+and+Yejin+Choi%2E+2019%2E+HellaSwag%3A+Can+a+machine+really+finish+your+sentence%3F+In+Proceedings+of+the+57th+Annual+Meeting+of+the+Association+for+Com)
40. Liyi Zhang, Jake Snell, and Thomas L. Griffiths. 2025a. Amortized bayesian meta-learning for low-rank adaptation of large language models. Conference on Language Modeling (COLM). [find](https://scholar.google.com/scholar?q=Liyi+Zhang%2C+Jake+Snell%2C+and+Thomas+L%2E+Griffiths%2E+2025a%2E+Amortized+bayesian+meta%2Dlearning+for+low%2Drank+adaptation+of+large+language+models%2E+Conference+on+Language+Modeling+%28COLM%29%2E)
41. Qiyuan Zhang, Fuyuan Lyu, Zexu Sun, Lei Wang, Weixu Zhang, Wenyue Hua, Haolun Wu, Zhihan Guo, Yufei Wang, Niklas Muennighoff, and 1 others. 2025b. A survey on test-time scaling in large language models: What, how, where, and how well? arXiv preprint arXiv:2503.24235. [find](https://scholar.google.com/scholar?q=Qiyuan+Zhang%2C+Fuyuan+Lyu%2C+Zexu+Sun%2C+Lei+Wang%2C+Weixu+Zhang%2C+Wenyue+Hua%2C+Haolun+Wu%2C+Zhihan+Guo%2C+Yufei+Wang%2C+Niklas+Muennighoff%2C+and+1+others%2E+2025b%2E+A+survey+on+test%2Dtime+scaling+in+large+language+model)

*Updated 1 November 2026*