Zhiyu Li

Zhiyu Li

Auto-generated profile, based on publicly available information

Zhiyu Li

MemTensor (Shanghai) Technology Co Ltd

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 November 2023 in Electronics
Verified
Journal Article
Open access
PDF file
ZLZhiyu Li
Zhiyu Li
YCYanfang Chen
Yanfang Chen
XZXuan Zhang
Xuan Zhang
LXLiang Xun
Liang Xun
With the continuous development and change exhibited by large language model (LLM) technology, represented by generative pretrained transformers (GPTs), many classic scenarios in various fields have re-emerged with new opportunities. This paper takes ChatGPT as the modeling object, incorporates LLM technology into the typical book resource understanding and recommendation scenario for the first time, and puts it into practice. By building a ChatGPT-like book recommendation system (BookGPT) framework based on ChatGPT, this paper attempts to apply ChatGPT to recommendation modeling for three typical tasks: book rating recommendation, user rating recommendation, and the book summary recommendation; it also explores the feasibility of LLM technology in book recommendation scenarios. At the same time, based on different evaluation schemes for book recommendation tasks and the existing classic recommendation models, this paper discusses the advantages and disadvantages of the BookGPT in book recommendation scenarios and analyzes the opportunities and improvement directions for subsequent LLMs in these scenarios. The experimental research shows the following: (1) The BookGPT can achieve good recommendation results in existing classic book recommendation tasks. Especially in cases containing less information about the target object to be recommended, such as zero-shot or one-shot learning tasks, the performance of the BookGPT is close to or even better than that of the current classic book recommendation algorithms, and this method has great potential for improvement. (2) In text generation tasks such as book summary recommendation, the recommendation effect of the BookGPT model is better than that of the manual editing process of Douban Reading, and it can even perform personalized interpretable content recommendations based on readers’ attribute and identity information, making it more persuasive than interpretable one-size-fits-all recommendation models. Finally, we have open-sourced the relevant datasets and experimental codes, hoping that the exploratory program proposed in this paper can inspire the development of more LLMs to expand their applications and theoretical research prospects in the field of book recommendation and general recommendation tasks.

6 February 2025 in Patterns
Verified
Journal Article
Open access
ZZZifan Zheng
Zifan Zheng
YWYezhaohui Wang
Yezhaohui Wang
YHYuxin Huang
Yuxin Huang
SSShichao Song
Shichao Song
MYMingchuan Yang
Mingchuan Yang
ZLZhiyu Li
Zhiyu Li
Large language models (LLMs) have demonstrated performance approaching human levels in tasks such as long-text comprehension and mathematical reasoning, but they remain black-box systems. Understanding the reasoning bottlenecks of LLMs remains a critical challenge, as these limitations are deeply tied to their internal architecture. Attention heads play a pivotal role in reasoning and are thought to share similarities with human brain functions. In this review, we explore the roles and mechanisms of attention heads to help demystify the internal reasoning processes of LLMs. We first introduce a four-stage framework inspired by the human thought process. Using this framework, we review existing research to identify and categorize the functions of specific attention heads. Additionally, we analyze the experimental methodologies used to discover these special heads and further summarize relevant evaluation methods and benchmarks. Finally, we discuss the limitations of current research and propose several potential future directions.

1 January 2024
Verified
Conference Paper
Open access
PDF file
XLXun Liang
Xun Liang
SSShichao Song
Shichao Song
SNSimin Niu
Simin Niu
ZLZhiyu Li
Zhiyu Li
FXFeiyu Xiong
Feiyu Xiong
HDHaiying Deng
Haiying Deng
Xun Liang, Shichao Song, Simin Niu, Zhiyu Li, Feiyu Xiong, Bo Tang, Yezhaohui Wang, Dawei He, Cheng Peng, Zhonghao Wang, Haiying Deng. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

1 January 2025
Verified
Conference Paper
Open access
PDF file
JZJihao Zhao
Jihao Zhao
ZJZhiyuan Ji
Zhiyuan Ji
ZFZhaoxin Fan
Zhaoxin Fan
HWHanyu Wang
Hanyu Wang
SNSimin Niu
Simin Niu
ZLZhiyu Li
Zhiyu Li
Jihao Zhao, Zhiyuan Ji, Zhaoxin Fan, Hanyu Wang, Simin Niu, Bo Tang, Feiyu Xiong, Zhiyu Li. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

28 September 2018 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
AZAakas Zhiyuli
Aakas Zhiyuli
LXLiang Xun
Liang Xun
YCYanfang Chen
Yanfang Chen
XDXiaoyong Du
Xiaoyong Du
Given the edge list of a social network, the node embedding method learns the structural features for every node and embeds the features into a vector space. The current related work on node embedding exploits only a portion of existing networks, e.g., static networks. However, social networks are inherently hierarchical and dynamic systems in which the topology changes constantly and the strength of influence of information among neighbors varies with different numbers of hops. We propose a highly efficient node embedding method, DNPS, that is faster and more accurate than state-of-the-art methods and that can further boost the training progress, especially under dynamic conditions. In this paper, we attempt to model the hierarchical and dynamic features of social networks by designing a damping-based sampling algorithm corresponding to a local search-based incremental learning algorithm, which can easily be extended to large-scale scenarios. We conduct extensive experiments on six real-world social networks with three challenging tasks, including missing link prediction, dynamic link prediction and multi-label classification. The results of the experiments on these tasks demonstrate that the proposed method significantly outperforms the existing methods with different settings.

1 January 2025
Verified
Conference Paper
Open access
PDF file
XWXunzhi Wang
Xunzhi Wang
ZZZhuowei Zhang
Zhuowei Zhang
GCGaonan Chen
Gaonan Chen
QLQiongyu Li
Qiongyu Li
BLBitong Luo
Bitong Luo
MHMengting Hu
Mengting Hu
Xunzhi Wang, Zhuowei Zhang, Gaonan Chen, Qiongyu Li, Bitong Luo, Zhixin Han, Haotian Wang, Zhiyu Li, Hang Gao, Mengting Hu. Findings of the Association for Computational Linguistics: ACL 2025. 2025.