Publications

Manuscripts

• Nuclear Deployed: Analyzing Catastrophic Risks in Decision-making of Autonomous LLM Agents
  Rongwu Xu*, Xiaojian Li*, Shuo Chen*, Wei Xu
  arXiv Preprint
  [Paper][Project Page][Code][X Post][ai safety China]
  
• Humanity's Last Exam
  Long Phan, Alice Gatti, Ziwen Han, Nathaniel Li, ... Rongwu Xu ..., Summer Yue, Alexandr Wang, Dan Hendrycks
  arXiv Preprint
  [Paper][Project Page]
  
• Rules Created by Symbolic Systems Cannot Constrain a Learning System
  Shih-Wai Lin, Rongwu Xu, Xiaojian Li, Wei Xu
  SSRN Preprint
  [Paper]
  
• DebateQA: Evaluating Question Answering on Debatable Knowledge
  Rongwu Xu*, Xuan Qi*, Zehan Qi, Wei Xu, Zhijiang Guo
  arXiv Preprint
  [Paper][Code]
  

2025

• On the Role of Attention Heads in Large Language Model Safety
  Zhenhong Zhou, Haiyang Yu, Xinghua Zhang, Rongwu Xu, Fei Huang, Kun Wang, Yang Liu, Junfeng Fang, Yongbin Li
  ICLR 2025 Oral [CORE A*]
  [Paper][Code]
  

2024

• Course-Correction: Safety Alignment Using Synthetic Preferences
  Rongwu Xu*, Yishuo Cai*, Zhenhong Zhou, Renjie Gu, Haiqin Wang, Yan Liu, Tianwei Zhang, Wei Xu, Han Qiu
  EMNLP 2024 [CORE A*]
  [Paper][Code][Poster][X Post]
  
• Knowledge Conflicts for LLMs: A Survey
  Rongwu Xu*, Zehan Qi*, Zhijiang Guo, Cunxiang Wang, Hongru Wang, Yue Zhang, Wei Xu
  EMNLP 2024 [CORE A*]
  [Paper][Code][机器之心][Talk (Chinese)][Slides][Poster][X Post]
  
• $LONG^{2}RAG$: Evaluating Long-Context & Long-Form Retrieval-Augmented Generation with Key Point Recall
  Zehan Qi*, Rongwu Xu*, Zhijiang Guo, Cunxiang Wang, Hao Zhang, Wei Xu
  EMNLP 2024 Findings [CORE A*]
  [Paper]
  
• Walking in Others' Shoes: How Perspective-Taking Guides Large Language Models in Reducing Toxicity and Bias
  Rongwu Xu, Zi'an Zhou, Tianwei Zhang, Zehan Qi, Su Yao, Ke Xu, Wei Xu, Han Qiu
  EMNLP 2024 [CORE A*]
  [Paper][Poster]
  
• Sing it, Narrate it: Quality Musical Lyrics Translation
  Zhuorui Ye, Jinhan Li, Rongwu Xu^
  EMNLP 2024 Findings [CORE A*]
  [Paper]
  
• How Alignment and Jailbreak Work: Explain LLM Safety through Intermediate Hidden States
  Zhenhong Zhou, Haiyang Yu, Xinghua Zhang, Rongwu Xu, Fei Huang, Yongbin Li
  EMNLP 2024 Findings [CORE A*]
  [Paper][Code][Poster]
  
• MR-BEN: A Comprehensive Meta-Reasoning Benchmark for Large Language Models
  Zhongshen Zeng, Yinhong Liu, Yingjia Wan, Jingyao Li, Pengguang Chen, Jianbo Dai, Yuxuan Yao, Rongwu Xu, Zehan Qi, Wanru Zhao, Linling Shen, Jianqiao Lu, Haochen Tan, Yukang Chen, Hao Zhang, Zhan Shi, Bailin Wang, Zhijiang Guo, Jiaya Jia
  NeurIPS 2024 [CORE A*]
  [Paper][Project Page][Code]
  
• Preemptive Answer ``Attacks'' on Chain-of-Thought Reasoning
  Rongwu Xu*, Zehan Qi*, Wei Xu
  ACL 2024 Findings [CORE A*]
  [Paper][Code][Poster]
  
• The Earth is Flat because...: Investigating LLMs' Belief towards Misinformation via Persuasive Conversation
  Rongwu Xu, Brian S. Lin, Shujian Yang, Tianqi Zhang, Weiyan Shi, Tianwei Zhang, Zhixuan Fang, Wei Xu, Han Qiu
  ACL 2024 Oral [CORE A*]
  🏆 Outstanding Paper Award [Certificate]
  [Paper][Project Page][Code][机器之心][Video][Poster]
  
• Exploring Chinese Humor Generation: A Study on Two-Part Allegorical Sayings
  Rongwu Xu
  IJCNN 2024 [CORE B]
  [Paper]
  
• Tempo: Confidentiality Preservation in Cloud-Based Neural Network Training
  Rongwu Xu and Zhixuan Fang
  IJCNN 2024 [CORE B]
  [Paper]
  
• LSync: A Universal Timeline-synchronizing Solution for Live Streaming
  Fan Dang*, Yifan Xu*, Rongwu Xu, Xinlei Chen, Yunhao Liu
  IEEE/ACM Trans. on Networking [JCR Q2]
  [Paper]
  

2023

• MISO: Legacy-compatible Privacy-preserving Single Sign-on using Trusted Execution Environments
  Rongwu Xu, Sen Yang, Fan Zhang, Zhixuan Fang
  EuroS&P 2023 [CORE A]
  [Paper][Project Page]
  

2022

• LSync: A Universal Event-synchronizing Solution for Live Streaming
  Yifan Xu, Fan Dang, Rongwu Xu, Xinlei Chen, Yunhao Liu
  INFOCOM 2022 [CORE A*]
  [Paper]
  
• LifeRec: A Mobile App for Lifelog Recording and Ubiquitous Recommendation
  Jiayu Li, Hantian Zhang*, Zhiyu He*, Rongwu Xu*, Pingfei Wu*, Min Zhang, Yiqun Liu, Shaoping Ma
  CHIIR 2022
  [Paper][Code]
  

* Equal Contribution, ^ Advising Role