Publications

2026

1. NeurIPS

   CMPQ: Compensatory Quantization via Input-Aware Hessian Damping

   Deng Pan, Grigorii Khvatskii, Xiaobao Huang, and 4 more authors

   In Advances in Neural Information Processing Systems, 2026

   Bib HTML

   @inproceedings{pan2026cmpq,
     title = {CMPQ: Compensatory Quantization via Input-Aware Hessian Damping},
     author = {Pan, Deng and Khvatskii, Grigorii and Huang, Xiaobao and Li, Peiyu and Gao, Shangqian and Hua, Ting and Chawla, Nitesh V.},
     booktitle = {Advances in Neural Information Processing Systems},
     year = {2026},
   }

2. NeurIPS

   PRO: Enabling Precise and Robust Text Watermark for Open-Source LLMs

   Jiaqi Xue, Yifei Zhao, Mansour Al Ghanim, and 4 more authors

   In Advances in Neural Information Processing Systems, 2026

   Bib HTML

   @inproceedings{xue2026pro,
     title = {PRO: Enabling Precise and Robust Text Watermark for Open-Source LLMs},
     author = {Xue, Jiaqi and Zhao, Yifei and Ghanim, Mansour Al and Gao, Shangqian and Sun, Ruimin and Lou, Qian and Zheng, Mengxin},
     booktitle = {Advances in Neural Information Processing Systems},
     year = {2026},
   }

3. EMNLP

   SALT: Salience-Aware Lexical Trie for Long-Context Compression

   Oteo Mamo, Hyunjin Yi, Joydhriti Choudhury, and 2 more authors

   In Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2026

   Bib HTML

   @inproceedings{mamo2026salt,
     title = {SALT: Salience-Aware Lexical Trie for Long-Context Compression},
     author = {Mamo, Oteo and Yi, Hyunjin and Choudhury, Joydhriti and Gao, Shangqian and Yu, Weikuan},
     booktitle = {Proceedings of the Conference on Empirical Methods in Natural Language Processing},
     year = {2026},
   }

4. EMNLP

   GeoMotionGPT: Geometry-Aligned Motion Understanding with Large Language Models

   Zhankai Ye, Bofan Li, Yukai Jin, and 5 more authors

   In Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2026

   Bib HTML

   @inproceedings{ye2026geomotiongpt,
     title = {GeoMotionGPT: Geometry-Aligned Motion Understanding with Large Language Models},
     author = {Ye, Zhankai and Li, Bofan and Jin, Yukai and Li, Shuoqiu and Wang, Wei and Zhang, Yanfu and Gao, Shangqian and Liu, Xin},
     booktitle = {Proceedings of the Conference on Empirical Methods in Natural Language Processing},
     year = {2026},
   }

5. EMNLP

   Enabling Confidence-Aware Control of Effective Memorization in Large Language Models

   Chenjie Ni, Omid Mersa, Zhepeng Wang, and 4 more authors

   In Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2026

   Bib HTML

   @inproceedings{ni2026confidence,
     title = {Enabling Confidence-Aware Control of Effective Memorization in Large Language Models},
     author = {Ni, Chenjie and Mersa, Omid and Wang, Zhepeng and Bao, Runxue and Li, Qun and Gao, Shangqian and Zhang, Yanfu},
     booktitle = {Proceedings of the Conference on Empirical Methods in Natural Language Processing},
     year = {2026},
   }

6. ACL Findings

   HyperEdit: Unlocking Instruction-based Text Editing in LLMs via Hypernetworks

   Yiming Zeng, Jinghan Cao, Zexin Li, and 7 more authors

   In Findings of the Association for Computational Linguistics: ACL 2026, 2026

   Bib HTML

   @inproceedings{zeng2026hyperedit,
     title = {HyperEdit: Unlocking Instruction-based Text Editing in LLMs via Hypernetworks},
     author = {Zeng, Yiming and Cao, Jinghan and Li, Zexin and Yu, Wanhao and Ye, Zhankai and Xiang, Dawei and Hua, Ting and Liu, Xin and Gao, Shangqian and Yu, Tingting},
     booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
     year = {2026},
   }

7. AAAI

   Query-Efficient Domain Knowledge Stealing Against Large Language Models

   Zhengao Li, Xiaopeng Yuan, Bolin Shen, and 5 more authors

   In Proceedings of the AAAI Conference on Artificial Intelligence, 2026

   Bib HTML

   @inproceedings{li2026knowledge,
     title = {Query-Efficient Domain Knowledge Stealing Against Large Language Models},
     author = {Li, Zhengao and Yuan, Xiaopeng and Shen, Bolin and Le, Kien and Wang, Haohan and Zhou, Xugui and Gao, Shangqian and Dong, Yushun},
     booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
     year = {2026},
   }

8. TMLR

   ToMoE: Converting Dense Large Language Models to Mixture-of-Experts through Dynamic Structural Pruning

   Shangqian Gao, Ting Hua, Reza Shirkavand, and 10 more authors

   Transactions on Machine Learning Research, 2026

   Presented at ICML 2026 (journal-to-conference track).

   Bib HTML

   @article{gao2026tomoe,
     title = {ToMoE: Converting Dense Large Language Models to Mixture-of-Experts through Dynamic Structural Pruning},
     author = {Gao, Shangqian and Hua, Ting and Shirkavand, Reza and Lin, Chi-Heng and Tang, Zheng and Li, Zhengao and Yuan, Longge and Li, Fangyi and Zhang, Zeyu and Ganjdanesh, Alireza and Lou, Qian and Xu, Jie and Hsu, Yen-Chang},
     journal = {Transactions on Machine Learning Research},
     year = {2026},
     note = {Presented at ICML 2026 (journal-to-conference track).},
   }

2025

1. NeurIPS

   ALTER: All-in-One Layer Pruning and Temporal Expert Routing for Efficient Diffusion Generation

   Xiaomeng Yang, Lei Lu, Qihui Fan, and 5 more authors

   In Advances in Neural Information Processing Systems, 2025

   Bib HTML

   @inproceedings{yang2025alter,
     title = {ALTER: All-in-One Layer Pruning and Temporal Expert Routing for Efficient Diffusion Generation},
     author = {Yang, Xiaomeng and Lu, Lei and Fan, Qihui and Yang, Changdi and Lin, Juyi and Wang, Yanzhi and Zhang, Xuan and Gao, Shangqian},
     booktitle = {Advances in Neural Information Processing Systems},
     year = {2025},
   }

2. NeurIPS

   Cost-Aware Contrastive Routing for LLMs

   Reza Shirkavand, Shangqian Gao, Peiran Yu, and 1 more author

   In Advances in Neural Information Processing Systems, 2025

   Spotlight

   Bib HTML

   @inproceedings{shirkavand2025routing,
     title = {Cost-Aware Contrastive Routing for LLMs},
     author = {Shirkavand, Reza and Gao, Shangqian and Yu, Peiran and Huang, Heng},
     booktitle = {Advances in Neural Information Processing Systems},
     year = {2025},
     note = {Spotlight},
   }

3. NeurIPS

   Any Large Language Model Can Be a Reliable Judge: Debiasing with a Reasoning-based Bias Detector

   Haoyan Yang, Runxue Bao, Cao Xiao, and 4 more authors

   In Advances in Neural Information Processing Systems, 2025

   Bib HTML

   @inproceedings{yang2025judge,
     title = {Any Large Language Model Can Be a Reliable Judge: Debiasing with a Reasoning-based Bias Detector},
     author = {Yang, Haoyan and Bao, Runxue and Xiao, Cao and Ma, Jun and Bhatia, Parminder and Gao, Shangqian and Kass-Hout, Taha},
     booktitle = {Advances in Neural Information Processing Systems},
     year = {2025},
   }

4. NeurIPS

   DictPFL: Efficient and Private Federated Learning on Encrypted Gradients

   Jiaqi Xue, Mayank Kumar, Yuzhang Shang, and 5 more authors

   In Advances in Neural Information Processing Systems, 2025

   Bib HTML

   @inproceedings{xue2025dictpfl,
     title = {DictPFL: Efficient and Private Federated Learning on Encrypted Gradients},
     author = {Xue, Jiaqi and Kumar, Mayank and Shang, Yuzhang and Gao, Shangqian and Ning, Rui and Zheng, Mengxin and Jiang, Xiaoqian and Lou, Qian},
     booktitle = {Advances in Neural Information Processing Systems},
     year = {2025},
   }

5. EMNLP

   Controllable Memorization in LLMs via Weight Pruning

   Chenjie Ni, Zhepeng Wang, Runxue Bao, and 2 more authors

   In Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2025

   Bib HTML

   @inproceedings{ni2025memorization,
     title = {Controllable Memorization in LLMs via Weight Pruning},
     author = {Ni, Chenjie and Wang, Zhepeng and Bao, Runxue and Gao, Shangqian and Zhang, Yanfu},
     booktitle = {Proceedings of the Conference on Empirical Methods in Natural Language Processing},
     year = {2025},
   }

6. CVPR

   Efficient Fine-Tuning and Concept Suppression for Pruned Diffusion Models

   Reza Shirkavand, Peiran Yu, Shangqian Gao, and 3 more authors

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2025

   Bib HTML

   @inproceedings{shirkavand2025suppression,
     title = {Efficient Fine-Tuning and Concept Suppression for Pruned Diffusion Models},
     author = {Shirkavand, Reza and Yu, Peiran and Gao, Shangqian and Somepalli, Gowthami and Goldstein, Tom and Huang, Heng},
     booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
     year = {2025},
   }

7. NAACL

   FlexiGPT: Pruning and Extending Large Language Models with Low-Rank Weight Sharing

   James Seale Smith, Chi-Heng Lin, Shikhar Tuli, and 5 more authors

   In The 2025 Annual Conference of the Nations of the Americas Chapter of the ACL, 2025

   Bib HTML

   @inproceedings{anonymous2025flexigpt,
     title = {Flexi{GPT}: Pruning and Extending Large Language Models with Low-Rank Weight Sharing},
     author = {Smith, James Seale and Lin, Chi-Heng and Tuli, Shikhar and Jeelani, Haris and Gao, Shangqian and Shen, Yilin and Jin, Hongxia and Hsu, Yen-Chang},
     booktitle = {The 2025 Annual Conference of the Nations of the Americas Chapter of the ACL},
     year = {2025},
     url = {https://openreview.net/forum?id=XDrzeWCePD}
   }

8. ICLR

   MoDeGPT: Modular Decomposition for Large Language Model Compression

   Chi-Heng Lin, Shangqian Gao, James Seale Smith, and 5 more authors

   In The Thirteenth International Conference on Learning Representations, 2025

   Oral

   Bib HTML

   @inproceedings{iclr2025modegpt,
     note = {Oral},
     title = {MoDe{GPT}: Modular Decomposition for Large Language Model Compression},
     author = {Lin, Chi-Heng and Gao, Shangqian and Smith, James Seale and Patel, Abhishek and Tuli, Shikhar and Shen, Yilin and Jin, Hongxia and Hsu, Yen-Chang},
     booktitle = {The Thirteenth International Conference on Learning Representations},
     year = {2025},
     url = {https://openreview.net/forum?id=8EfxjTCg2k}
   }

9. ICLR

   Not All Prompts Are Made Equal: Prompt-based Pruning of Text-to-Image Diffusion Models

   Alireza Ganjdanesh, Reza Shirkavand, Shangqian Gao, and 1 more author

   In The Thirteenth International Conference on Learning Representations, 2025

   Bib HTML

   @inproceedings{iclr2025not,
     title = {Not All Prompts Are Made Equal: Prompt-based Pruning of Text-to-Image Diffusion Models},
     author = {Ganjdanesh, Alireza and Shirkavand, Reza and Gao, Shangqian and Huang, Heng},
     booktitle = {The Thirteenth International Conference on Learning Representations},
     year = {2025},
     url = {https://openreview.net/forum?id=3BhZCfJ73Y}
   }

2024

1. NeurIPS

   DISP-LLM: Dimension-Independent Structural Pruning for Large Language Models

   Shangqian Gao, Chi-Heng Lin, Ting Hua, and 4 more authors

   In Advances in Neural Information Processing Systems (NeurIPS), 2024

   Bib HTML

   @inproceedings{neurips24,
     author = {Gao, Shangqian and Lin, Chi-Heng and Hua, Ting and Tang, Zheng and Shen, Yilin and Jin, Hongxia and Hsu, Yen-Chang},
     title = {DISP-LLM: Dimension-Independent Structural Pruning for Large Language Models},
     booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
     year = {2024},
   }

2. EMNLP

   Unlocking Memorization in Large Language Models with Dynamic Soft Prompting

   Zhepeng Wang, Runxue Bao, Yawen Wu, and 6 more authors

   In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), 2024

   Bib HTML

   @inproceedings{emnlp24,
     author = {Wang, Zhepeng and Bao, Runxue and Wu, Yawen and Taylor, Jackson and Xiao, Cao and Zheng, Feng and Jiang, Weiwen and Gao*, Shangqian and Zhang*, Yanfu},
     title = {Unlocking Memorization in Large Language Models with Dynamic Soft Prompting},
     booktitle = {Proceedings of the Conference on Empirical Methods in Natural
                       Language Processing (EMNLP)},
     year = {2024},
   }

3. NAACL

   Adaptive Rank Selections for Low-Rank Approximation of Language Models

   Shangqian Gao, Ting Hua, Yen-Chang Hsu, and 2 more authors

   In Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL), 2024

   Bib HTML

   @inproceedings{naacl24,
     author = {Gao, Shangqian and Hua, Ting and Hsu, Yen-Chang and Shen, Yilin and Jin, Hongxia},
     title = {Adaptive Rank Selections for Low-Rank Approximation of Language Models},
     booktitle = {Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL)},
     year = {2024},
   }

4. CVPR

   Auto-Train-Once: Controller Network Guided Automatic Network Pruning from Scratch

   Xidong Wu*, Shangqian Gao*, Zeyu Zhang, and 5 more authors

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024

   Bib HTML

   @inproceedings{cvpr24_1,
     author = {Wu*, Xidong and Gao*, Shangqian and Zhang, Zeyu and Li, Zhenzhen and Bao, Runxue and Zhang, Yanfu and Wang, Xiaoqian and Huang, Heng},
     title = {Auto-Train-Once: Controller Network Guided Automatic Network Pruning from Scratch},
     booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
     year = {2024},
   }

5. CVPR

   Device-Wise Federated Network Pruning

   Shangqian Gao*, Junyi Li*, Zeyu Zhang*, and 3 more authors

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024

   Bib HTML

   @inproceedings{cvpr24_2,
     author = {Gao*, Shangqian and Li*, Junyi and Zhang*, Zeyu and Zhang, Yanfu and Cai, Weidong and Huang, Heng},
     title = {Device-Wise Federated Network Pruning},
     booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
     year = {2024},
   }

6. CVPR

   Jointly Training and Pruning CNNs via Learnable Agent Guidance and Alignment

   Alireza Ganjdanesh*, Shangqian Gao*, and Heng Huang

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024

   Bib HTML

   @inproceedings{cvpr24_3,
     author = {Ganjdanesh*, Alireza and Gao*, Shangqian and Huang, Heng},
     title = {Jointly Training and Pruning CNNs via Learnable Agent Guidance and Alignment},
     booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
     year = {2024},
   }

7. CVPR

   BilevelPruning: Unified Dynamic and Static Channel Pruning for Convolutional Neural Networks

   Shangqian Gao, Yanfu Zhang, Feihu Huang, and 1 more author

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024

   Bib HTML

   @inproceedings{cvpr24_4,
     author = {Gao, Shangqian and Zhang, Yanfu and Huang, Feihu and Huang, Heng},
     title = {BilevelPruning: Unified Dynamic and Static Channel Pruning for Convolutional Neural Networks},
     booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
     year = {2024},
   }

8. AAAI

   Compressing Image-to-Image Translation GANs Using Local Density Structures on Their Learned Manifold

   Alireza Ganjdanesh*, Shangqian Gao*, and Heng Huang

   In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), 2024

   Bib HTML

   @inproceedings{ali2024gan,
     title = {Compressing Image-to-Image Translation GANs Using Local Density Structures on Their Learned Manifold},
     author = {Ganjdanesh*, Alireza and Gao*, Shangqian and Huang, Heng},
     booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence (AAAI)},
     year = {2024},
   }

9. WACV

   Token Fusion: Bridging the Gap between Token Pruning and Token Merging

   Minchul Kim, Shangqian Gao, Yen-Chang Hsu, and 2 more authors

   In Winter Conference on Computer Vision (WACV), 2024

   Bib HTML

   @inproceedings{wacv24,
     author = {Kim, Minchul and Gao, Shangqian and Hsu, Yen{-}Chang and Shen, Yilin and Jin, Hongxia},
     title = {Token Fusion: Bridging the Gap between Token Pruning and Token Merging},
     booktitle = {Winter Conference on Computer Vision (WACV)},
     year = {2024},
   }

2023

1. ICLR

   Learning to Jointly Share and Prune Weights for Grounding Based Vision and Language Models

   Shangqian Gao, Burak Uzkent, Yilin Shen, and 2 more authors

   In International Conference on Learning Representations (ICLR), 2023
2. AAAI

   EffConv: Efficient Learning of Kernel Sizes for Convolution Layers of CNNs

   Alireza Ganjdanesh, Shangqian Gao, and Heng Huang

   In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), 2023
3. ICCV

   Structural alignment for network pruning through partial regularization

   Shangqian Gao, Zeyu Zhang, Yanfu Zhang, and 2 more authors

   Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023
4. TPAMI

   Gradient Descent Ascent for Minimax Problems on Riemannian Manifolds

   Feihu Huang, and Shangqian Gao

   IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2023
5. EMNLP

   Dynamic Low-rank Estimation for Transformer-based Language Models

   Ting Hua*, Xiao Li*, Shangqian Gao, and 3 more authors

   In Proceedings of the Conference on Empirical Methods in Natural Language Processing Findings (EMNLP findings), 2023

2022

1. ICLR

   Bregman Gradient Policy Optimization

   Feihu Huang*, Shangqian Gao*, and Heng Huang

   In International Conference on Learning Representations (ICLR), 2022
2. ECCV

   Disentangled differentiable network pruning

   Shangqian Gao, Feihu Huang, Yanfu Zhang, and 1 more author

   In European Conference on Computer Vision (ECCV), 2022
3. ECCV

   Interpretations steered network pruning via amortized inferred saliency maps

   Alireza Ganjdanesh*, Shangqian Gao*, and Heng Huang

   In European Conference on Computer Vision (ECCV), 2022
4. ECCV

   Recover fair deep classification models via altering pre-trained structure

   Yanfu Zhang, Shangqian Gao, and Heng Huang

   In European Conference on Computer Vision (ECCV), 2022
5. NN

   Riemannian gradient methods for stochastic composition problems

   Feihu Huang, and Shangqian Gao

   Neural Networks, 2022
6. JMLR

   Accelerated zeroth-order and first-order momentum methods from mini to minimax optimization

   Feihu Huang, Shangqian Gao, Jian Pei, and 1 more author

   The Journal of Machine Learning Research (JMLR), 2022
7. KDD

   Improving social network embedding via new second-order continuous graph neural networks

   Yanfu Zhang, Shangqian Gao, Jian Pei, and 1 more author

   In Proceedings of the 28th ACM SIGKDD conference on knowledge discovery and data mining (KDD), 2022
8. NeurIPS

   Enhanced bilevel optimization via bregman distance

   Feihu Huang, Junyi Li, Shangqian Gao, and 1 more author

   Advances in Neural Information Processing Systems, 2022

2021

1. CVPR

   Network pruning via performance maximization

   Shangqian Gao, Feihu Huang, Weidong Cai, and 1 more author

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021
2. ICCV

   Exploration and estimation for model compression

   Yanfu Zhang*, Shangqian Gao*, and Heng Huang

   In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021
3. ICCV

   Adversarial attack on deep cross-modal hamming retrieval

   Chao Li*, Shangqian Gao*, Cheng Deng, and 2 more authors

   In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021
4. JMLR

   Black-box reductions for zeroth-order gradient algorithms to achieve lower query complexity

   Bin Gu, Xiyuan Wei, Shangqian Gao, and 3 more authors

   The Journal of Machine Learning Research (JMLR), 2021

2020

1. CVPR

   Discrete model compression with resource constraint for deep neural networks

   Shangqian Gao, Feihu Huang, Jian Pei, and 1 more author

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020
2. ICML

   Momentum-based policy gradient methods

   Feihu Huang, Shangqian Gao, Jian Pei, and 1 more author

   In International conference on machine learning (ICML), 2020

2019

1. CVPR

   Cross Domain Model Compression by Structurally Weight Sharing

   Shangqian Gao, Cheng Deng, and Heng Huang

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Jun 2019
2. NeurIPS

   Cross-modal learning with adversarial samples

   Chao Li, Shangqian Gao, Cheng Deng, and 2 more authors

   Advances in neural information processing systems (NeurIPS), Jun 2019
3. IJCAI

   Zeroth-order stochastic alternating direction method of multipliers for nonconvex nonsmooth optimization

   Feihu Huang, Shangqian Gao, Songcan Chen, and 1 more author

   International Joint Conferences on Artificial Intelligence (IJCAI), Jun 2019

2018

1. AAAI

   Action prediction from videos via memorizing hard-to-predict samples

   Shangqian Gao*, Yu Kong*, Bin Sun, and 1 more author

   In Proceedings of the AAAI conference on artificial intelligence (AAAI), Jun 2018

2017

1. AAAI

   Video recovery via learning variation and consistency of images

   Zhouyuan Huo, Shangqian Gao, Weidong Cai, and 1 more author

   In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), Jun 2017

2016

1. TMM

   Discriminative multi-instance multitask learning for 3D action recognition

   Yanhua Yang, Cheng Deng, Shangqian Gao, and 3 more authors

   IEEE Transactions on Multimedia, Jun 2016