Elamur Research⌕
☰
home / papers / 2026 / Not All Directions Matter: Towards Structured and Task-Aware Low-Rank Model Adaptation

Not All Directions Matter: Towards Structured and Task-Aware Low-Rank Model Adaptation

Paper · 2026 · Xi Xiao, Chenrui Ma, Yunbei Zhang et al.

Summary

The paper argues that LoRA has two weaknesses. First, it spends its small rank budget equally on every low-rank direction, even though many directions carry noise; the authors call this semantic drift. Second, it tunes each layer in isolation, so updates in neighbouring layers can point in unrelated directions; they call this structural incoherence. Their method, StructLoRA, adds two components that are active only during training. A learnable gate, trained with an Information Bottleneck objective, scales each rank-one direction so that capacity goes to label-relevant directions. A shallow GNN (GCN or GAT) then treats layers as graph nodes and passes messages between their filtered updates, which brings the updates of related layers into alignment. After training, the coordinator is dropped and the update is merged into the base weights, so inference costs the same as plain LoRA. Experiments cover language, vision and multimodal models. StructLoRA beats LoRA, AdaLoRA, DoRA and sparsity- or rank-allocation methods, with the largest gains in low-rank and few-data settings. This makes it useful when compute and data are limited.

Key points

  • Motivating evidence: with standard LoRA, update gradients of adjacent layers showed low cosine similarity of only about 0.27–0.41.
  • Models evaluated: LLaMA, LLaVA, ViT (including ViT-B/16 on ImageNet), Qwen and Gemma 2. Baselines included LoRA, AdaLoRA, DoRA and dynamic-rank and pruning variants.
  • Filtering ablation: the IB-guided filter clearly outperformed Random Masking and Top-k Norm selection. This suggests that the size of an update direction says little about its relevance to the task.
  • Training overhead on LLaMA-7B: epochs were roughly 4–6% slower and peak GPU memory rose by under 0.8 GB. Inference latency is unchanged.
  • Qualitative analyses: Grad-CAM maps showed more focused attention on the relevant object. Layer-wise similarity heatmaps showed a block-diagonal pattern, meaning groups of related layers were coordinated.

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

🎓Harvard UniversityCambridge · United States112 works · 42,023 citations🎓Brown UniversityProvidence · United States76 works · 25,952 citations🎓Yale UniversityNew Haven · United States95 works · 12,765 citations🎓University of VirginiaCharlottesville · United States36 works · 9,730 citations🎓Northeastern UniversityBoston · United States57 works · 6,533 citations🎓University of Alabama at BirminghamBirmingham · United States6 works · 9,739 citations🎓University of BristolBristol · United Kingdom14 works · 725 citations

Also on the byline: Tulane University.

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 1Research subjects and languages the work is about.Efficient models & compression · 242 works

Cite

Cite as Xi Xiao, Chenrui Ma, Yunbei Zhang, Chen Liu, Zhuxuanzi Wang, Yanshu Li, Lin Zhao, Guosheng Hu, Tianyang Wang and Hao Xu. 2026. Not All Directions Matter: Towards Structured and Task-Aware Low-Rank Model Adaptation. 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.97

@inproceedings{xiao2026not,
  title = {{Not All Directions Matter: Towards Structured and Task-Aware Low-Rank Model Adaptation}},
  author = {Xi Xiao, Chenrui Ma, Yunbei Zhang, Chen Liu, Zhuxuanzi Wang, Yanshu Li, Lin Zhao, Guosheng Hu, Tianyang Wang and Hao Xu},
  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.97},
  url = {https://doi.org/10.18653/v1/2026.acl-long.97}
}

References

94 references, 25 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Alexander A. Alemi, Ian Fischer, Joshua V. Dillon, and Kevin Murphy. 2017. Deep variational information bottleneck. In International Conference on Learning Representations (ICLR). find
  2. Mohamed Ishmael Belghazi, Aristide Baratin, Sai Rajeshwar, Sherjil Ozair, Yoshua Bengio, Aaron Courville, and Devon Hjelm. 2018. Mutual information neural estimation. In Proceedings of the 35th International Conference on Machine Learning (ICML), volume 80 of Proceedings of Machine Learning Research… find
  3. Elad Ben-Zaken, Yoav Goldberg, and Shauli Ravfogel. 2022. Bitfit: Simple parameter-efficient fine-tuning for transformer-based masked language models. In Findings of the Association for Computational Linguistics: ACL 2022, pages 1–9.
  4. Yonatan Bisk, Rowan Zellers, Jianfeng Gao, Yejin Choi, and 1 others. 2020. Piqa: Reasoning about physical commonsense in natural language. In Proceedings of the AAAI conference on artificial intelligence, volume 34, pages 7432–7439.
  5. Dorian Drost. 2024. An overview of the lora family. Medium. find
  6. Samuel Lima Braz. 2025. Peft: Parameter-efficient fine-tuning methods for llms. Hugging Face Blog. find
  7. Zhen Fang, Zhuoyang Liu, Jiaming Liu, Hao Chen, Yu Zeng, Shiting Huang, Zehui Chen, Lin Chen, Shanghang Zhang, and Feng Zhao. 2025. Dualvla: Building a generalizable embodied agent via partial decoupling of reasoning and action. arXiv preprint arXiv:2511.22134. — [arXiv:2511.22134](https://arxiv.org/abs/2511.22134) find
  8. Simon Carbonnelle and Christophe De Vleeschouwer. 2018. Layer rotation: a surprisingly powerful indicator of generalization in deep networks? arXiv preprint arXiv:1806.01603. — [arXiv:1806.01603](https://arxiv.org/abs/1806.01603) find
  9. Zhiwei Chen, Yupeng Hu, Zhiheng Fu, Zixu Li, Jiale Huang, Qinlei Huang, and Yinwei Wei. 2026. Intent: Invariance and discrimination-aware noise mitigation for robust composed image retrieval. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 20463–20471. find
  10. Zhiwei Chen, Yupeng Hu, Zixu Li, Zhiheng Fu, Xuemeng Song, and Liqiang Nie. 2025. Offset: Segmentation-based focus shift revision for composed image retrieval. In Proceedings of the ACM International Conference on Multimedia, page 6113–6122. find
  11. 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. In Proceedings of the 2019 conference of the north American chapter of the association for computational linguistics:…
  12. 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. arXiv preprint arXiv:1803.05457.
  13. Tim Dettmers, Artidoro Pagnoni, Ari Holtzman, and Luke Zettlemoyer. 2023. Qlora: Efficient finetuning of quantized llms. In Advances in Neural Information Processing Systems (NeurIPS).
  14. Haonan Dong, Kehan Jiang, Haoran Ye, Wenhao Zhu, Zhaolu Kang, and Guojie Song. 2026. Neureasoner: Towards explainable, controllable, and unified reasoning via mixture-of-neurons. Preprint, arXiv:2604.02972.
  15. Haonan Dong, Wenhao Zhu, Guojie Song, and Liang Wang. 2025. AuroRA: Breaking low-rank bottleneck of loRA with nonlinear mapping. In The Thirty-ninth Annual Conference on Neural Information Processing Systems. find
  16. Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. 2021. An image is worth 16×16 words: Transformers for image recognition at scale. In Intern… find
  17. Gemma Team. 2024. Gemma 2: Improving open language models at a practical size. arXiv preprint arXiv:2408.00118. find
  18. Ruiyan Han, Zhen Fang, XinYu Sun, Yuchen Ma, Ziheng Wang, Yu Zeng, Zehui Chen, Lin Chen, Wenxuan Huang, Wei-Jie Xu, and 1 others. 2026. Unicorn: Towards self-improving unified multimodal models through self-generated supervision. arXiv preprint arXiv:2601.03193.
  19. Zeyu Han, Chao Gao, Jinyang Liu, Jeff Zhang, and Sai Qian Zhang. 2024. Parameter-efficient fine-tuning for large models: A comprehensive survey. arXiv preprint arXiv:2403.14608.
  20. Junxian He, Chunting Zhou, Xuezhe Ma, Taylor Berg-Kirkpatrick, and Graham Neubig. 2022. Towards a unified view of parameter-efficient transfer learning. In International Conference on Learning Representations (ICLR).
  21. Neil Houlsby, Andrei Giurgiu, Stanislaw Jastrzebski, Bruna Morrone, Quentin de Laroussilhe, Andrea Gesmundo, Mona Attariyan, and Sylvain Gelly. 2019. Parameter-efficient transfer learning for nlp. arXiv preprint arXiv:1902.00751. find
  22. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022a. Lora: Low-rank adaptation of large language models. In International Conference on Learning Representations (ICLR).
  23. Jinyi Hu, Xiaoyuan Yi, Wenhao Li, Maosong Sun, and Xing Xie. 2022b. Fuse it more deeply! a variational transformer with layer-wise latent variable inference for text generation. In Proceedings of NAACL-HLT 2022, pages 697–716. find
  24. Yupeng Hu, Zixu Li, Zhiwei Chen, Qinlei Huang, Zhiheng Fu, Mingzhu Xu, and Liqiang Nie. 2026. Refine: Composed video retrieval via shared and differential semantics enhancement. ACM Transactions on Multimedia Computing, Communications and Applications. find
  25. Yuxuan Hu, Jing Zhang, Zhe Zhao, Cuiping Li, and Hong Chen. 2024. Sp-lora: Sparsity-preserved low-rank adaptation for sparse large language models. find
  26. OpenReview (ICLR 2025; withdrawn submission). find
  27. Yuxuan Hu, Jing Zhang, Zhe Zhao, Cuiping Li, and Hong Chen. 2025. Lors: Efficient low-rank adaptation for sparse large language model. arXiv preprint arXiv:2501.08582. — [arXiv:2501.08582](https://arxiv.org/abs/2501.08582) find
  28. Eric Jang, Shixiang Gu, and Ben Poole. 2017. Categorical reparameterization with gumbel-softmax. In International Conference on Learning Representations (ICLR).
  29. Jiachen Jiang, Jinxin Zhou, and Zhihui Zhu. 2024. Tracing representation progression: Analyzing and enhancing layer-wise similarity. arXiv preprint arXiv:2406.14479. — [arXiv:2406.14479](https://arxiv.org/abs/2406.14479) find
  30. Kehan Jiang, Haonan Dong, Zhaolu Kang, Zhengzhou Zhu, and Guojie Song. 2026. Foe: Forest of errors makes the first solution the best in large reasoning models. Preprint, arXiv:2604.02967.
  31. Thomas N. Kipf and Max Welling. 2017. Semi-supervised classification with graph convolutional networks. In International Conference on Learning Representations (ICLR).
  32. Dawid J. Kopiczko, Tijmen Blankevoort, and Yuki M. Asano. 2024. Vera: Vector-based random matrix adaptation. In International Conference on Learning Representations (ICLR).
  33. Ariel N. Lee, Cole J. Hunter, and Nataniel Ruiz. 2023. Platypus: Quick, cheap, and powerful refinement of llms. arXiv preprint arXiv:2308.07317. find
  34. Junxian Li, Kai Liu, Leyang Chen, Weida Wang, Zhixin Wang, Jiaqi Xu, Fan Li, Renjing Pei, Linghe Kong, and Yulun Zhang. 2026a. Planviz: Evaluating planning-oriented image generation and editing for computer-use tasks. arXiv preprint arXiv:2602.06663. — [arXiv:2602.06663](https://arxiv.org/abs/2602.06663) find
  35. Junxian Li, Xinyue Xu, Sai Ma, and Sichao Li. 2025. Faithact: Faithfulness planning and acting in mllms. arXiv preprint arXiv:2511.08409. — [arXiv:2511.08409](https://arxiv.org/abs/2511.08409) find
  36. Xiang Lisa Li and Percy Liang. 2021. Prefix-tuning: Optimizing continuous prompts for generation. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics (ACL), pages 4582–4597. find
  37. Zixu Li, Yupeng Hu, Zhiwei Chen, Qinlei Huang, Guozhi Qiu, Zhiheng Fu, and Meng Liu. 2026b. Retrack: Evidence-driven dual-stream directional anchor calibration network for composed video retrieval. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 23373– 23381. find
  38. Zixu Li, Yupeng Hu, Zhiwei Chen, Shiqi Zhang, Qinlei Huang, Zhiheng Fu, and Yinwei Wei. 2026c. Habit: Chrono-synergia robust progressive learning framework for composed image retrieval. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 6762–6770. find
  39. Danqi Liao, Chen Liu, Benjamin W Christensen, Alexander Tong, Guillaume Huguet, Guy Wolf, Maximilian Nickel, Ian Adelstein, and Smita Krishnaswamy. 2024. Assessing neural network representations during training using noise-resilient diffusion spectral entropy. In 2024 58th Annual Conference on Infor… find
  40. Yang Lin, Xinyu Ma, Xu Chu, Yujie Jin, Zhibang Yang, Yasha Wang, and Hong Mei. 2024. Lora dropout as a sparsity regularizer for overfitting control. Preprint, arXiv:2404.09610. — [arXiv:2404.09610](https://arxiv.org/abs/2404.09610) find
  41. Chen Liu, Xingzhi Sun, Xi Xiao, Alexandre Van Tassel, Ke Xu, Kristof Reimann, Danqi Liao, Mark Gerstein, Tianyang Wang, Xiao Wang, and Smita Krishnaswamy. 2026. Dispersion loss counteracts embedding condensation and improves generalization in small language models. arXiv preprint arXiv:2602.00217. — [arXiv:2602.00217](https://arxiv.org/abs/2602.00217) find
  42. Haotian Liu, Chunyuan Li, Yuheng Li, and Yong Jae Lee. 2024a. Improved baselines with visual instruction tuning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
  43. Shih-Yang Liu, Chien-Yi Wang, Hongxu Yin, Pavlo Molchanov, Yu-Chiang Frank Wang, Kwang-Ting Cheng, and Min-Hung Chen. 2024b. DoRA: Weight-decomposed low-rank adaptation. In Proceedings of the 41st International Conference on Machine Learning (ICML), volume 235 of Proceedings of Machine Learning Rese… find
  44. Yiyang Liu, James Chenhao Liang, Heng Fan, Wenhao Yang, Yiming Cui, Xiaotian Han, Lifu Huang, Dongfang Liu, Qifan Wang, and Cheng Han. All you need is one: Capsule prompt tuning with a single vector. In The Thirty-ninth Annual Conference on Neural Information Processing Systems. find
  45. Man Luo, Shuguang Chen, and Chitta Baral. 2022. A simple approach to jointly rank passages and select relevant sentences in the obqa context. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Student Res… find
  46. Yuren Mao, Yuhang Ge, Yijiang Fan, Wenyi Xu, Yu Mi, Zhonghao Hu, and Yunjun Gao. 2025. A survey on lora of large language models. Frontiers of Computer Science, 19(7):197605. find
  47. Meta AI. 2024. Introducing llama 3.1: Our most capable models to date. Blog post. find
  48. Jonas Pfeiffer, Aishwarya Kamath, Andreas Rücklé, Kyunghyun Cho, and Iryna Gurevych. 2021. Adapterfusion: Non-destructive task composition for transfer learning. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL), pages 487–503.
  49. Nusrat Jahan Prottasha, Upama Roy Chowdhury, Shetu Mohanto, Tasfia Nuzhat, Abdullah As Sami, Md Shamol Ali, Md Shohanur Islam Sobuj, Hafijur Raman, Md Kowsher, and Ozlem Ozmen Garibay. 2025. Peft a2z: Parameter-efficient fine-tuning survey for large language and vision models. arXiv preprint arXiv:2… — [arXiv:2504.14117](https://arxiv.org/abs/2504.14117) find
  50. Guozhi Qiu, Zhiwei Chen, Zixu Li, Qinlei Huang, Zhiheng Fu, Xuemeng Song, and Yupeng Hu. 2026. Melt: Improve composed image retrieval via the modification frequentation-rarity balance network. arXiv preprint arXiv:2603.29291. — [arXiv:2603.29291](https://arxiv.org/abs/2603.29291) find
  51. Naftali Tishby and Noga Zaslavsky. 2015. Deep learning and the information bottleneck principle. In 2015 IEEE Information Theory Workshop (ITW), pages 1–5. IEEE. find
  52. Zeju Qiu, Yutong Liu, Weijia Wen, Li Xu, Yixiao Zhang, Jian Yang, Tong Lu, and Ying Shan. 2023. Controlling text-to-image diffusion by orthogonal finetuning. In Advances in Neural Information Processing Systems (NeurIPS). find
  53. Maithra Raghu, Thomas Unterthiner, Simon Kornblith, Chiyuan Zhang, and Alexey Dosovitskiy. 2021. Do vision transformers see like convolutional neural networks? In Advances in Neural Information Processing Systems (NeurIPS). find
  54. Adithya Renduchintala, Tugrul Konuk, and Oleksii Kuchaiev. 2024. Tied-lora: Enhancing parameter efficiency of lora with weight tying. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), pages 8… find
  55. Keisuke Sakaguchi, Ronan Le Bras, Chandra Bhagavatula, and Yejin Choi. 2021. Winogrande: An adversarial winograd schema challenge at scale. Communications of the ACM, 64(9):99–106. find
  56. Karthik Abinav Sankararaman, Soham De, Zheng Xu, W Ronny Huang, and Tom Goldstein. 2020. The impact of neural network overparameterization on gradient confusion and stochastic gradient descent. In International conference on machine learning, pages 8469–8479. PMLR. find
  57. Ramprasaath R. Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, and Dhruv Batra. 2017. Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE International Conference on Computer Vision (ICCV), pages 618–626.
  58. Hanwen Shen, Ting Ying, Jiajie Lu, and Shanshan Wang. 2026. Preconditioned test-time adaptation for out-of-distribution debiasing in narrative generation. arXiv preprint arXiv:2603.13683.
  59. H. U. K. Shinwari and 1 others. 2025. Ard-lora: Dynamic rank allocation for parameter-efficient fine-tuning of large models. arXiv preprint arXiv:2506.18267. — [arXiv:2506.18267](https://arxiv.org/abs/2506.18267) find
  60. Alon Talmor, Jonathan Herzig, Nicholas Lourie, and Jonathan Berant. 2019. Commonsenseqa: A question answering challenge targeting commonsense Hugo Touvron, Matthieu Cord, Alaaeldin El-Nouby, Jakob Verbeek, and Hervé Jégou. 2022. Three things everyone should know about vision transformers. In Compute… find
  61. Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Gabriel Rémi, Naman Goyal, Eric Hambro, Faisal Azhar, and 1 others. 2023. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971.
  62. Mojtaba Valipour, Mehdi Rezagholizadeh, Ivan Kobyzev, and Ali Ghodsi. 2023. Dylora: Parameter-efficient tuning of pretrained models using dynamic search-free low-rank adaptation. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics (EACL), pa… find
  63. Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Liò, and Yoshua Bengio. 2018. Graph attention networks. In International Conference on Learning Representations (ICLR). find
  64. Alex Wang, Amanpreet Singh, Julian Michael, Felix Hill, Omer Levy, and Samuel R. Bowman. 2018. Glue: A multi-task benchmark and analysis platform for natural language understanding. In Proceedings of the 2018 EMNLP Workshop BlackboxNLP, pages 353–355.
  65. Janet Wang, Yunbei Zhang, Zhengming Ding, and Jihun Hamm. 2025a. Doctor approved: Generating medically accurate skin disease images through ai-expert feedback. In Advances in Neural Information Processing Systems (NeurIPS). find
  66. Zhuxuanzi Wang, Mingqiao Mo, Xi Xiao, Chen Liu, Chenrui Ma, Yunbei Zhang, Xiao Wang, Smita Krishnaswamy, and Tianyang Wang. 2025b. Ctr-lora: Curvature-aware and trust-region guided low-rank adaptation for large language models. arXiv preprint arXiv:2510.15962. — [arXiv:2510.15962](https://arxiv.org/abs/2510.15962) find
  67. Xi Xiao, Aristeidis Tsaris, Anika Tabassum, John Lagergren, Larry M York, Tianyang Wang, and Xiao Wang. 2025a. Focus: Fused observation of channels for unveiling spectra. arXiv preprint arXiv:2507.14787. — [arXiv:2507.14787](https://arxiv.org/abs/2507.14787) find
  68. Xi Xiao, Wentao Wang, Jiacheng Xie, Lijing Zhu, Gaofei Chen, Zhengji Li, Tianyang Wang, and Min Xu. 2024. Hgtdp-dta: Hybrid graph-transformer with dynamic prompt for drug-target binding affinity prediction. In International Conference on Neural Information Processing, pages 340–354. Springer. find
  69. Xi Xiao, Zhuxuanzi Wang, Mingqiao Mo, Chen Liu, Chenrui Ma, Yanshu Li, Smita Krishnaswamy, Xiao Wang, and Tianyang Wang. 2026a. Self-supervised visual prompting for cross-domain road damage detection. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 3514–352… find
  70. Yu Zeng, Wenxuan Huang, Zhen Fang, Shuang Chen, Yufan Shen, Yishuo Cai, Xiaoman Wang, Zhenfei Yin, Lin Chen, Zehui Chen, and 1 others. 2026. Vision-deepresearch benchmark: Rethinking visual and textual search for multimodal large language models. arXiv preprint arXiv:2602.02185. — [arXiv:2602.02185](https://arxiv.org/abs/2602.02185) find
  71. Hao Zhang, Bo Huang, Zhenjia Li, Xi Xiao, Hui Yi Leong, Zumeng Zhang, Xinwei Long, Tianyang Wang, and Hao Xu. 2025a. Sensitivity-lora: Low-load sensitivity-based fine-tuning for large language models. arXiv preprint arXiv:2509.09119. find
  72. Xi Xiao, Yunbei Zhang, Xingjian Li, Tianyang Wang, Xiao Wang, Yuxiang Wei, Jihun Hamm, and Min Xu. 2025b. Visual instance-aware prompt tuning. In Proceedings of the 33rd ACM International Conference on Multimedia, pages 2880–2889. find
  73. Hao Zhang, Zhenjia Li, Runfeng Bao, Yifan Gao, Xi Xiao, Heng Zhang, Shuyang Zhang, Bo Huang, Yuhang Wu, Tianyang Wang, and 1 others. 2025b. Hyperadalora: Accelerating lora rank allocation during training via hypernetworks without sacrificing performance. arXiv preprint arXiv:2510.02630.
  74. Xi Xiao, Yunbei Zhang, Yanshuh Li, Xingjian Li, Tianyang Wang, Jihun Hamm, Xiao Wang, and Min Xu. 2025c. Visual variational autoencoder prompt tuning. arXiv preprint arXiv:2503.17650. — [arXiv:2503.17650](https://arxiv.org/abs/2503.17650) find
  75. Mingyang Zhang, Hao Chen, Chunhua Shen, Zhen Yang, Linlin Ou, Xinyi Yu, and Bohan Zhuang. 2023a. Loraprune: Structured pruning meets low-rank parameter-efficient fine-tuning. arXiv preprint arXiv:2305.18403. find
  76. Xi Xiao, Yunbei Zhang, Lin Zhao, Yiyang Liu, Xiaoying Liao, Zheda Mai, Xingjian Li, Xiao Wang, Hao Xu, Jihun Hamm, Xue Lin, Min Xu, Qifan Wang, Tianyang Wang, and Cheng Han. 2026b. Prompt-based adaptation in large-scale vision models: A survey. Transactions on Machine Learning Research. find
  77. Yi Xin, Siqi Luo, Haodi Zhou, Junlong Du, Xiaohong Liu, Yue Fan, Qing Li, and Yuntao Du. 2024. Parameter-efficient fine-tuning for pre-trained vision models: A survey. arXiv preprint arXiv:2402.02242. — [arXiv:2402.02242](https://arxiv.org/abs/2402.02242) find
  78. Zhangchen Xu, Fengqing Jiang, Luyao Niu, Yuntian Deng, Radha Poovendran, Yejin Choi, and Bill Yuchen Lin. 2024. Magpie: Alignment data synthesis from scratch by prompting aligned llms with nothing. Preprint, arXiv:2406.08464.
  79. An Yang, Beichen Zhang, Binyuan Hui, Yang Fan, Xingzhang Ren, Jingren Zhou, Junyang Lin, and 1 others. 2024. Qwen2.5 technical report. arXiv preprint arXiv:2412.15115. find
  80. Qianyun Yang, Zhiwei Chen, Yupeng Hu, Zixu Li, Zhiheng Fu, and Liqiang Nie. 2026. Stable: Efficient hybrid nearest neighbor search via magnitude-uniformity and cardinality-robustness. arXiv preprint arXiv:2604.01617. — [arXiv:2604.01617](https://arxiv.org/abs/2604.01617) find
  81. Shuo Yuan, Yifan Sun, Ming Jiang, Xinyu Zhang, Zhipeng Liu, Zhiqiang Li, and Ming-Hsuan Xu. 2024. Householder reflection adaptation: Bridging the gap between low-rank and orthogonal adaptation. arXiv preprint arXiv:2405.17484. — [arXiv:2405.17484](https://arxiv.org/abs/2405.17484) find
  82. 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. find
  83. Mingyu Zhang, Zixu Li, Zhiwei Chen, Zhiheng Fu, Xiaowei Zhu, Jiajia Nie, Yinwei Wei, and Yupeng Hu. 2026a. Hint: Composed image retrieval with dual-path compositional contextualized network. arXiv preprint arXiv:2603.26341. — [arXiv:2603.26341](https://arxiv.org/abs/2603.26341) find
  84. Qianchi Zhang, Hainan Zhang, Liang Pang, Hongwei Zheng, and Zhiming Zheng. 2026b. Stable-rag: Mitigating retrieval-permutation-induced hallucinations in retrieval-augmented generation. arXiv preprint arXiv:2601.02993.
  85. Qingru Zhang, Minshuo Chen, Alexander Bukharin, Pengcheng He, Yu Cheng, Weizhu Chen, and Tuo Zhao. 2023b. Adaptive budget allocation for parameter-efficient fine-tuning. In International Conference on Learning Representations (ICLR). find
  86. Xueyan Zhang, Jinman Zhao, Zhifei Yang, Yibo Zhong, Shuhao Guan, Linbo Cao, and Yining Wang. 2025c. Uora: Uniform orthogonal reinitialization adaptation in parameter efficient fine-tuning of large models. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Vol…
  87. Yunbei Zhang, Chengyi Cai, Feng Liu, and Jihun Hamm. 2025d. Prime Once, then Reprogram Locally: An Efficient Alternative to Black-Box Model Reprogramming. OpenReview. find
  88. Yunbei Zhang, Akshay Mehra, and Jihun Hamm. 2025e. Ot-vp: Optimal transport-guided visual prompting for test-time adaptation. In IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 1122–1132. IEEE. Oral Presentation. find
  89. Yunbei Zhang, Akshay Mehra, Shuaicheng Niu, and Jihun Hamm. 2025f. Dpcore: Dynamic prompt coreset for continual test-time adaptation. In Forty-second International Conference on Machine Learning (ICML). find
  90. Yunbei Zhang, Shuaicheng Niu, Feng Liu, and Jihun Hamm. 2026c. Adapting in the Dark: Towards Stable and Efficient Black-Box Test-Time Adaptation. OpenReview. find
  91. Jinman Zhao, Xueyan Zhang, Jiaru Li, Jingcheng Niu, Yulan Hu, Erxue Min, and Gerald Penn. 2025a. Tiny budgets, big gains: Parameter placement strategy in parameter super-efficient fine-tuning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 6326–6344.
  92. Lin Zhao, Xinru Jiang, Xi Xiao, Qihui Fan, Lei Lu, Yanzhi Wang, Xue Lin, Octavia Camps, Pu Zhao, and Jianyang Gu. 2026. Hieramp: Coarse-to-fine autoregressive amplification for generative dataset distillation. arXiv preprint arXiv:2603.06932. — [arXiv:2603.06932](https://arxiv.org/abs/2603.06932) find
  93. Lin Zhao, Yushu Wu, Xinru Jiang, Jianyang Gu, Yanzhi Wang, Xiaolin Xu, Pu Zhao, and Xue Lin. 2025b. Taming diffusion for dataset distillation with high representativeness. arXiv preprint arXiv:2505.18399. — [arXiv:2505.18399](https://arxiv.org/abs/2505.18399) find
  94. Yixiao Zhou, Yang Li, Dongzhou Cheng, Hehe Fan, and Yu Cheng. 2026. Look inward to explore outward: Learning temperature policy from llm internal states via hierarchical rl. arXiv preprint arXiv:2602.13035. — [arXiv:2602.13035](https://arxiv.org/abs/2602.13035) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/xiao-2026-not-all-directions-matter-structured/