Training-Trajectory-Aware Token Selection.

Zeyu Qin

Zhanming Shen Jiaqi Hu Zeyu Qin Hao Chen 0081 Wentao Ye Zenan Huang Yihong Zhuang Guoshan Lu Jun Zhou 0011 Junbo Zhao 0001 Training-Trajectory-Aware Token Selection. 2026 January abs/2601.10348 CoRR https://doi.org/10.48550/arXiv.2601.10348 db/journals/corr/corr2601.html#abs-2601-10348 streams/journals/corr

Zhanming Shen Zeyu Qin Jiaqi Hu Wentao Ye Hao Chen 0081 Xiaomeng Hu Haokai Xu Gang Chen 0001 Yi R. Fung 0001 Haobo Wang 0001 Supervised Fine-Tuning Needs to Unlock the Potential of Token Priority. 2026 February abs/2602.01227 CoRR https://doi.org/10.48550/arXiv.2602.01227 db/journals/corr/corr2602.html#abs-2602-01227 streams/journals/corr

Ziniu Li Congliang Chen Tian Xu 0003 Zeyu Qin Jiancong Xiao Zhi-Quan Luo Ruoyu Sun 0001 Preserving Diversity in Supervised Fine-Tuning of Large Language Models. 2025 ICLR https://openreview.net/forum?id=NQEe7B7bSw conf/iclr/2025 db/conf/iclr/iclr2025.html#LiC0QXL025 Haoyu Wang 0018 Zeyu Qin Li Shen 0008 Xueqian Wang 0001 Dacheng Tao Minhao Cheng Safety Reasoning with Guidelines. 2025 ICML https://proceedings.mlr.press/v267/wang25cg.html https://openreview.net/forum?id=BHwWLeXDYF conf/icml/2025 db/conf/icml/icml2025.html#WangQ00TC25

Haoyu Wang 0018 Zeyu Qin Li Shen 0008 Xueqian Wang 0001 Minhao Cheng Dacheng Tao Leveraging Reasoning with Guidelines to Elicit and Utilize Knowledge for Enhancing Safety Alignment. 2025 February abs/2502.04040 CoRR https://doi.org/10.48550/arXiv.2502.04040 db/journals/corr/corr2502.html#abs-2502-04040 streams/journals/corr

Zeyu Qin Qingxiu Dong Xingxing Zhang 0002 Li Dong 0004 Xiaolong Huang Ziyi Yang 0011 Mahmoud Khademi Dongdong Zhang 0001 Hany Hassan Awadalla Yi R. Fung 0001 Weizhu Chen Minhao Cheng Furu Wei Scaling Laws of Synthetic Data for Language Models. 2025 March abs/2503.19551 CoRR https://doi.org/10.48550/arXiv.2503.19551 db/journals/corr/corr2503.html#abs-2503-19551 streams/journals/corr

Hongzheng Yang Yongqiang Chen 0002 Zeyu Qin Tongliang Liu Chaowei Xiao Kun Zhang 0001 Bo Han 0003 Does Representation Intervention Really Identify Desired Concepts and Elicit Alignment? 2025 May abs/2505.18672 CoRR https://doi.org/10.48550/arXiv.2505.18672 db/journals/corr/corr2505.html#abs-2505-18672 streams/journals/corr

Haoyu Wang 0018 Zeyu Qin Yifei Zhao Chao Du Min Lin Xueqian Wang 0001 Tianyu Pang Lifelong Safety Alignment for Language Models. 2025 May abs/2505.20259 CoRR https://doi.org/10.48550/arXiv.2505.20259 db/journals/corr/corr2505.html#abs-2505-20259 streams/journals/corr

Ren-Jian Wang Ke Xue 0001 Zeyu Qin Ziniu Li Sheng Tang Hao-Tian Li Shengcai Liu Chao Qian 0001 Quality-Diversity Red-Teaming: Automated Generation of High-Quality and Diverse Attackers for Large Language Models. 2025 June abs/2506.07121 CoRR https://doi.org/10.48550/arXiv.2506.07121 db/journals/corr/corr2506.html#abs-2506-07121 streams/journals/corr

Zenan Huang Yihong Zhuang Guoshan Lu Zeyu Qin Haokai Xu Tianyu Zhao Ru Peng Jiaqi Hu Zhanming Shen Xiaomeng Hu Xijun Gu Peiyi Tu Jiaxin Liu Wenyu Chen Yuzhuo Fu Zhiting Fan Yanmei Gu Yuanyuan Wang Zhengkai Yang Jianguo Li Junbo Zhao 0002 Reinforcement Learning with Rubric Anchors. 2025 August abs/2508.12790 CoRR https://doi.org/10.48550/arXiv.2508.12790 db/journals/corr/corr2508.html#abs-2508-12790 streams/journals/corr

Zhanming Shen Zeyu Qin Zenan Huang Hao Chen 0081 Jiaqi Hu Yihong Zhuang Guoshan Lu Gang Chen 0001 Junbo Zhao 0002 Merge-of-Thought Distillation. 2025 September abs/2509.08814 CoRR https://doi.org/10.48550/arXiv.2509.08814 db/journals/corr/corr2509.html#abs-2509-08814 streams/journals/corr

Haotian Luo Huaisong Zhang Xuelin Zhang Haoyu Wang 0018 Zeyu Qin Wenjie Lu Guozheng Ma Haiying He Yingsha Xie Qiyang Zhou Zixuan Hu Hongze Mi Yibo Wang Naiqiang Tan Hong Chen Yi R. Fung 0001 Chun Yuan 0003 Li Shen 0008 UltraHorizon: Benchmarking Agent Capabilities in Ultra Long-Horizon Scenarios. 2025 September abs/2509.21766 CoRR https://doi.org/10.48550/arXiv.2509.21766 db/journals/corr/corr2509.html#abs-2509-21766 streams/journals/corr

Feng Ju Zeyu Qin Rui Min Zhitao He 0001 Lingpeng Kong Yi R. Fung 0001 Reasoning Path Divergence: A New Metric and Curation Strategy to Unlock LLM Diverse Thinking. 2025 October abs/2510.26122 CoRR https://doi.org/10.48550/arXiv.2510.26122 db/journals/corr/corr2510.html#abs-2510-26122 streams/journals/corr

Zhitao He 0001 Haolin Yang Zeyu Qin Yi R. Fung 0001 MedTutor-R1: Socratic Personalized Medical Teaching with Multi-Agent Simulation. 2025 December abs/2512.05671 CoRR https://doi.org/10.48550/arXiv.2512.05671 db/journals/corr/corr2512.html#abs-2512-05671 streams/journals/corr

Rui Min Zeyu Qin Nevin L. Zhang Li Shen 0008 Minhao Cheng Uncovering, Explaining, and Mitigating the Superficial Safety of Backdoor Defense. 2024 NeurIPS http://papers.nips.cc/paper_files/paper/2024/hash/8e8399e5e7aed601c9f135f40be26564-Abstract-Conference.html conf/nips/2024 db/conf/nips/neurips2024.html#MinQZ0C24

Haoyu Wang 0018 Guozheng Ma Ziqiao Meng Zeyu Qin Li Shen 0008 Zhong Zhang 0014 Bingzhe Wu Liu Liu 0014 Yatao Bian Tingyang Xu Xueqian Wang 0001 Peilin Zhao Step-On-Feet Tuning: Scaling Self-Alignment of LLMs via Bootstrapping. 2024 abs/2402.07610 CoRR https://doi.org/10.48550/arXiv.2402.07610 db/journals/corr/corr2402.html#abs-2402-07610

Yupeng Chen Senmiao Wang Zhihang Lin Zeyu Qin Yushun Zhang Tian Ding Ruoyu Sun 0001 MoFO: Momentum-Filtered Optimizer for Mitigating Forgetting in LLM Fine-Tuning. 2024 abs/2407.20999 CoRR https://doi.org/10.48550/arXiv.2407.20999 db/journals/corr/corr2407.html#abs-2407-20999 streams/journals/corr

Ziniu Li Congliang Chen Tian Xu 0003 Zeyu Qin Jiancong Xiao Ruoyu Sun 0001 Zhi-Quan Luo Entropic Distribution Matching in Supervised Fine-tuning of LLMs: Less Overfitting and Better Diversity. 2024 abs/2408.16673 CoRR https://doi.org/10.48550/arXiv.2408.16673 db/journals/corr/corr2408.html#abs-2408-16673 streams/journals/corr

Rui Min Zeyu Qin Nevin L. Zhang Li Shen 0008 Minhao Cheng Uncovering, Explaining, and Mitigating the Superficial Safety of Backdoor Defense. 2024 abs/2410.09838 CoRR https://doi.org/10.48550/arXiv.2410.09838 db/journals/corr/corr2410.html#abs-2410-09838 streams/journals/corr

Zeyu Qin Haipeng Yao Tianle Mai Di Wu 0001 Ni Zhang Song Guo 0001 Multi-Agent Reinforcement Learning Aided Computation Offloading in Aerial Computing for the Internet-of-Things. 1976-1986 2023 May - June 16 IEEE Trans. Serv. Comput. 3 https://doi.org/10.1109/TSC.2022.3190562 db/journals/tsc/tsc16.html#QinYMWZG23

Liang Chen 0001 Yang Deng 0002 Yatao Bian Zeyu Qin Bingzhe Wu Tat-Seng Chua Kam-Fai Wong Beyond Factuality: A Comprehensive Evaluation of Large Language Models as Knowledge Generators. 6325-6341 2023 EMNLP https://doi.org/10.18653/v1/2023.emnlp-main.390 https://aclanthology.org/2023.emnlp-main.390 conf/emnlp/2023 db/conf/emnlp/emnlp2023.html#ChenDBQWCW23 Zeyu Qin Liuyi Yao Daoyuan Chen Yaliang Li Bolin Ding Minhao Cheng Revisiting Personalized Federated Learning: Robustness Against Backdoor Attacks. 4743-4755 2023 KDD https://doi.org/10.1145/3580305.3599898 conf/kdd/2023 db/conf/kdd/kdd2023.html#QinYCLDC23 Ziniu Li Tian Xu 0003 Zeyu Qin Yang Yu 0001 Zhi-Quan Luo Imitation Learning from Imperfection: Theoretical Justifications and Algorithms. 2023 NeurIPS http://papers.nips.cc/paper_files/paper/2023/hash/3af25aa3de8b7b02ddbd1b6be5031be8-Abstract-Conference.html conf/nips/2023 db/conf/nips/neurips2023.html#LiXQ0L23 Rui Min Zeyu Qin Li Shen 0008 Minhao Cheng Towards Stable Backdoor Purification through Feature Shift Tuning. 2023 NeurIPS http://papers.nips.cc/paper_files/paper/2023/hash/ee37d51b3c003d89acba2363dde256af-Abstract-Conference.html conf/nips/2023 db/conf/nips/neurips2023.html#MinQ0C23

Zeyu Qin Liuyi Yao Daoyuan Chen Yaliang Li Bolin Ding Minhao Cheng Revisiting Personalized Federated Learning: Robustness Against Backdoor Attacks. 2023 abs/2302.01677 CoRR https://doi.org/10.48550/arXiv.2302.01677 db/journals/corr/corr2302.html#abs-2302-01677

Rui Min Zeyu Qin Li Shen 0008 Minhao Cheng Towards Stable Backdoor Purification through Feature Shift Tuning. 2023 abs/2310.01875 CoRR https://doi.org/10.48550/arXiv.2310.01875 db/journals/corr/corr2310.html#abs-2310-01875

Peiying Zhang 0001 Chao Wang 0093 Zeyu Qin Haotong Cao A multidomain virtual network embedding algorithm based on multiobjective optimization for Internet of Drones architecture in Industry 4.0. 710-728 2022 52 Softw. Pract. Exp. 3 https://doi.org/10.1002/spe.2815 db/journals/spe/spe52.html#ZhangWQC22

Zeyu Qin Yanbo Fan Yi Liu Li Shen 0008 Yong Zhang 0034 Jue Wang 0001 Baoyuan Wu Boosting the Transferability of Adversarial Attacks with Reverse Adversarial Perturbation. 2022 conf/nips/2022 NeurIPS http://papers.nips.cc/paper_files/paper/2022/hash/c0f9419caa85d7062c7e6d621a335726-Abstract-Conference.html db/conf/nips/neurips2022.html#QinFL0Z0W22

Peiying Zhang 0001 Chao Wang 0093 Zeyu Qin Haotong Cao A multi-domain VNE algorithm based on multi-objective optimization for IoD architecture in Industry 4.0. 2022 abs/2202.12830 CoRR https://arxiv.org/abs/2202.12830 db/journals/corr/corr2202.html#abs-2202-12830

Jiancong Xiao Zeyu Qin Yanbo Fan Baoyuan Wu Jue Wang 0001 Zhi-Quan Luo Adaptive Smoothness-weighted Adversarial Training for Multiple Perturbations with Its Stability Analysis. 2022 abs/2210.00557 CoRR https://doi.org/10.48550/arXiv.2210.00557 db/journals/corr/corr2210.html#abs-2210-00557

Zeyu Qin Yanbo Fan Yi Liu Li Shen 0008 Yong Zhang 0034 Jue Wang 0001 Baoyuan Wu Boosting the Transferability of Adversarial Attacks with Reverse Adversarial Perturbation. 2022 abs/2210.05968 CoRR https://doi.org/10.48550/arXiv.2210.05968 db/journals/corr/corr2210.html#abs-2210-05968

Zeyu Qin Liuqing Ye Zhe Wang Yue Qin Jiyuan Huang Xiong Xie Xiangwei Hu A Line Selection Method Based on Transient Phase Current Variation Comparison. 26:1-26:5 2021 ICITEE https://doi.org/10.1145/3513142.3513169 conf/icitee/2021 db/conf/icitee/icitee2021.html#QinYWQHXH21 Yutong Zhao Haipeng Yao Zeyu Qin Tianle Mai Collaborate Q-learning Aided Load Balance in Satellites Communications. 968-973 2021 IWCMC https://doi.org/10.1109/IWCMC51323.2021.9498837 conf/iwcmc/2021 db/conf/iwcmc/iwcmc2021.html#ZhaoYQM21 Zeyu Qin Yanbo Fan Hongyuan Zha Baoyuan Wu Random Noise Defense Against Query-Based Black-Box Attacks. 7650-7663 2021 NeurIPS https://proceedings.neurips.cc/paper/2021/hash/3eb414bf1c2a66a09c185d60553417b8-Abstract.html conf/nips/2021 db/conf/nips/neurips2021.html#QinFZW21

Zeyu Qin Yanbo Fan Hongyuan Zha Baoyuan Wu Theoretical Study of Random Noise Defense against Query-Based Black-Box Attacks. 2021 abs/2104.11470 CoRR https://arxiv.org/abs/2104.11470 db/journals/corr/corr2104.html#abs-2104-11470

Zeyu Qin Haipeng Yao Tianle Mai Traffic Optimization in Satellites Communications: A Multi-agent Reinforcement Learning Approach. 269-273 2020 IWCMC https://doi.org/10.1109/IWCMC48107.2020.9148523 conf/iwcmc/2020 db/conf/iwcmc/iwcmc2020.html#QinYM20 Yatao Bian Haotong Cao Congliang Chen Daoyuan Chen Gang Chen 0001 Hao Chen 0081 Hong Chen Liang Chen 0001 Weizhu Chen Wenyu Chen Yongqiang Chen 0002 Yupeng Chen Minhao Cheng Tat-Seng Chua Yang Deng 0002 Bolin Ding Tian Ding Li Dong 0004 Qingxiu Dong Chao Du Yanbo Fan Zhiting Fan Yuzhuo Fu Yi R. Fung 0001 Xijun Gu Yanmei Gu Song Guo 0001 Bo Han 0003 Hany HassanHany Hassan Awadalla Haiying He Zhitao He 0001 Jiaqi Hu Xiangwei Hu Xiaomeng Hu Zixuan Hu Jiyuan Huang Xiaolong Huang Zenan Huang Feng Ju Mahmoud Khademi Lingpeng Kong Hao-Tian Li Jianguo Li Yaliang Li Ziniu Li Min Lin Zhihang Lin Jiaxin Liu Liu Liu 0014 Shengcai Liu Tongliang Liu Yi Liu Guoshan Lu Wenjie Lu Haotian Luo Zhi-Quan Luo Guozheng Ma Tianle Mai Ziqiao Meng Hongze Mi Rui Min Tianyu Pang Ru Peng Chao Qian 0001 Yue Qin Li Shen 0008 Zhanming Shen Ruoyu Sun 0001 Naiqiang Tan Sheng Tang Dacheng Tao Peiyi Tu Chao Wang 0093 Haobo Wang 0001 Haoyu Wang 0018 Jue Wang 0001 Ren-Jian Wang Senmiao Wang Xueqian Wang 0001 Yibo Wang Yuanyuan Wang Zhe Wang Furu Wei Kam-Fai Wong Baoyuan Wu Bingzhe Wu Di Wu 0001 Chaowei Xiao Jiancong Xiao Xiong Xie Yingsha Xie Haokai Xu Tian Xu 0003 Tingyang Xu Ke Xue 0001 Haolin Yang Hongzheng Yang Zhengkai Yang Ziyi Yang 0011 Haipeng Yao Liuyi Yao Liuqing Ye Wentao Ye Yang Yu 0001 Chun Yuan 0003 Hongyuan Zha Dongdong Zhang 0001 Huaisong Zhang Kun Zhang 0001 Nevin Lianwen ZhangNevin L. Zhang Ni Zhang Peiying Zhang 0001 Xingxing Zhang 0002 Xuelin Zhang Yong Zhang 0034 Yushun Zhang Zhong Zhang 0014 Junbo Zhao 0001 Junbo Zhao 0002 Peilin Zhao Tianyu Zhao Yifei Zhao Yutong Zhao Jun Zhou 0011 Qiyang Zhou Yihong Zhuang