📝 Publications
* indicates equal contribution, and † indicates corresponding author.
MetaMem: Evolving Meta-Memory for Knowledge Utilization through Self-Reflective Symbolic Optimization
Haidong Xin$^{*}$, Xinze Li$^{*}$, Zhenghao Liu$^†$, Yukun Yan$^†$, Shuo Wang, Cheng Yang, Yu Gu, Ge Yu, Maosong Sun
This work proposes MetaMem, a meta-memory framework that evolves meta-memory through self-reflective symbolic optimization, enabling knowledge utilization in long-horizon human-LLM interactions.
Structured Knowledge Representation through Contextual Pages for Retrieval-Augmented Generation
Xinze Li, Zhenghao Liu$^†$, Haidong Xin, Yukun Yan$^†$, Shuo Wang, Zheni Zeng, Sen Mei, Ge Yu, Maosong Sun
This work enhances RAG systems by constructing structured cognitive outlines to guide iterative retrieval, effectively organizing multi-dimensional knowledge into coherent pages for more accurate answer generation.
LISRec: Modeling User Preferences with Learned Item Shortcuts for Sequential Recommendation
Haidong Xin, Zhenghao Liu$^†$, Sen Mei, Yukun Yan, Shi Yu, Shuo Wang, Zulong Chen, Yu Gu, Ge Yu, Chenyan Xiong
This work improves sequential recommendation systems by extracting personalized semantic shortcuts from user-item interaction histories, enhancing the capture of stable user preferences.
Knowledge Intensive Agents
Zhenghao Liu, Pengcheng Huang, Zhipeng Xu, Xinze Li, Shuliang Liu, Chunyi Peng, Haidong Xin, Yukun Yan$^†$, Shuo Wang, Xu Han, Zhiyuan Liu$^†$, Maosong Sun$^†$, Yu Gu, Ge Yu
This work provides a comprehensive overview of Retrieval-Augmented Generation from an agentic perspective, categorizing knowledge-intensive agents into acquisition and utilization roles, and highlighting future directions for joint optimization in multi-agent RAG systems.
Adapting Language Models to Text Matching based Recommendation Systems
Haidong Xin, Sen Mei, Zhenghao Liu$^†$, Xiaohua Li, Minghe Yu, Yu Gu, Ge Yu
This work has significantly improved the performance of text matching in recommendation systems by incorporating language model pretraining.
LLMsPark: A Benchmark for Evaluating Large Language Models in Strategic Gaming Contexts
Junhao Chen, Jingbo Sun, Xiang Li, Haidong Xin, Yuhao Xue, Yibin Xu, Hao Zhao$^†$
This work introduces LLMsPark, a game theory-based platform for evaluating large language models’ strategic behaviors and decision-making abilities in multi-agent environments, providing a novel criterion for assessing their intelligence.
MMAD: Multi-modal Movie Audio Description
Xiaojun Ye, Junhao Chen, Xiang Li, Haidong Xin, Chao Li, Sheng Zhou$^†$, Jiajun Bu
This work has unlocked a whole new experience of watching movies for the visually impaired.
Puzzle Game: Prediction and Classification of Wordle Solution Words
Haidong Xin$^{*†}$, Fang Wu$^{*}$, Zhitong Zhou$^{*}$
This work conducted a detailed numerical analysis of the Wordle game, revealing statistical patterns within it.