Zobrazeno 1 - 10
of 42
pro vyhledávání: '"Dou, Shihan"'
Autor:
Huang, Caishuang, Zhao, Wanxu, Zheng, Rui, Lv, Huijie, Dou, Shihan, Li, Sixian, Wang, Xiao, Zhou, Enyu, Ye, Junjie, Yang, Yuming, Gui, Tao, Zhang, Qi, Huang, Xuanjing
As the development of large language models (LLMs) rapidly advances, securing these models effectively without compromising their utility has become a pivotal area of research. However, current defense strategies against jailbreak attacks (i.e., effo
Externí odkaz:
http://arxiv.org/abs/2406.18118
Autor:
Bao, Rong, Zheng, Rui, Dou, Shihan, Wang, Xiao, Zhou, Enyu, Wang, Bo, Zhang, Qi, Ding, Liang, Tao, Dacheng
In aligning large language models (LLMs), utilizing feedback from existing advanced AI rather than humans is an important method to scale supervisory signals. However, it is highly challenging for AI to understand human intentions and societal values
Externí odkaz:
http://arxiv.org/abs/2406.11190
Autor:
Dou, Shihan, Liu, Yan, Zhou, Enyu, Li, Tianlong, Jia, Haoxiang, Xiong, Limao, Zhao, Xin, Ye, Junjie, Zheng, Rui, Gui, Tao, Zhang, Qi, Huang, Xuanjing
The success of Reinforcement Learning from Human Feedback (RLHF) in language model alignment is critically dependent on the capability of the reward model (RM). However, as the training process progresses, the output distribution of the policy model
Externí odkaz:
http://arxiv.org/abs/2405.00438
With the development of the open source community, the code is often copied, spread, and evolved in multiple software systems, which brings uncertainty and risk to the software system (e.g., bug propagation and copyright infringement). Therefore, it
Externí odkaz:
http://arxiv.org/abs/2405.00428
Autor:
Zhou, Weikang, Wang, Xiao, Xiong, Limao, Xia, Han, Gu, Yingshuang, Chai, Mingxu, Zhu, Fukang, Huang, Caishuang, Dou, Shihan, Xi, Zhiheng, Zheng, Rui, Gao, Songyang, Zou, Yicheng, Yan, Hang, Le, Yifan, Wang, Ruohui, Li, Lijun, Shao, Jing, Gui, Tao, Zhang, Qi, Huang, Xuanjing
Jailbreak attacks are crucial for identifying and mitigating the security vulnerabilities of Large Language Models (LLMs). They are designed to bypass safeguards and elicit prohibited outputs. However, due to significant differences among various jai
Externí odkaz:
http://arxiv.org/abs/2403.12171
Autor:
Lv, Huijie, Wang, Xiao, Zhang, Yuansen, Huang, Caishuang, Dou, Shihan, Ye, Junjie, Gui, Tao, Zhang, Qi, Huang, Xuanjing
Adversarial misuse, particularly through `jailbreaking' that circumvents a model's safety and ethical protocols, poses a significant challenge for Large Language Models (LLMs). This paper delves into the mechanisms behind such successful attacks, int
Externí odkaz:
http://arxiv.org/abs/2402.16717
Autor:
Xu, Nuo, Zhao, Jun, Zu, Can, Li, Sixian, Chen, Lu, Zhang, Zhihao, Zheng, Rui, Dou, Shihan, Qin, Wenjuan, Gui, Tao, Zhang, Qi, Huang, Xuanjing
Faithfulness, expressiveness, and elegance is the constant pursuit in machine translation. However, traditional metrics like \textit{BLEU} do not strictly align with human preference of translation quality. In this paper, we explore leveraging reinfo
Externí odkaz:
http://arxiv.org/abs/2402.11525
Autor:
Xi, Zhiheng, Chen, Wenxiang, Hong, Boyang, Jin, Senjie, Zheng, Rui, He, Wei, Ding, Yiwen, Liu, Shichun, Guo, Xin, Wang, Junzhe, Guo, Honglin, Shen, Wei, Fan, Xiaoran, Zhou, Yuhao, Dou, Shihan, Wang, Xiao, Zhang, Xinbo, Sun, Peng, Gui, Tao, Zhang, Qi, Huang, Xuanjing
In this paper, we propose R$^3$: Learning Reasoning through Reverse Curriculum Reinforcement Learning (RL), a novel method that employs only outcome supervision to achieve the benefits of process supervision for large language models. The core challe
Externí odkaz:
http://arxiv.org/abs/2402.05808
Autor:
Dou, Shihan, Liu, Yan, Jia, Haoxiang, Xiong, Limao, Zhou, Enyu, Shen, Wei, Shan, Junjie, Huang, Caishuang, Wang, Xiao, Fan, Xiaoran, Xi, Zhiheng, Zhou, Yuhao, Ji, Tao, Zheng, Rui, Zhang, Qi, Huang, Xuanjing, Gui, Tao
The advancement of large language models (LLMs) has significantly propelled the field of code generation. Previous work integrated reinforcement learning (RL) with compiler feedback for exploring the output space of LLMs to enhance code generation qu
Externí odkaz:
http://arxiv.org/abs/2402.01391
Autor:
Fan, Xiaoran, Ji, Tao, Jiang, Changhao, Li, Shuo, Jin, Senjie, Song, Sirui, Wang, Junke, Hong, Boyang, Chen, Lu, Zheng, Guodong, Zhang, Ming, Huang, Caishuang, Zheng, Rui, Xi, Zhiheng, Zhou, Yuhao, Dou, Shihan, Ye, Junjie, Yan, Hang, Gui, Tao, Zhang, Qi, Qiu, Xipeng, Huang, Xuanjing, Wu, Zuxuan, Jiang, Yu-Gang
Current large vision-language models (VLMs) often encounter challenges such as insufficient capabilities of a single visual component and excessively long visual tokens. These issues can limit the model's effectiveness in accurately interpreting comp
Externí odkaz:
http://arxiv.org/abs/2401.17221