Zeguan Xiao

Zeguan Xiao

Final-year PhD candidate @ SUFE SCAI. LLM Safety, Alignment, Robustness & Unlearning.

Homepage Corrections Design

Homepage Corrections Design

Date: 2026-04-19

Problem

The academic homepage at https://ZeguanXiao.github.io contains multiple factual errors across three areas: personal biography, CV page, and publication entries. Several publications point to wrong papers (incorrect titles, wrong authors, or arXiv IDs belonging to other researchers). The CV page still contains 100% template placeholder content.

Scope

Fix all factual errors. No structural changes to layouts, navigation, or styling. English-only for now (the /zh/ bilingual layer is a separate initiative).

Area 1: _pages/about.md

Three fixes:

  1. Chinese name typo — current: 肖泽冠; correct: 肖泽管
  2. School name wrong — current: School of Information Management and Engineering; correct: School of Computing and Artificial Intelligence (SCAI)
  3. Missing research interest — add LLM Unlearning to the bullet list

Area 2: _pages/cv.md

Full rewrite. Replace all GitHub University placeholder content with real information.

Education

DegreeInstitutionPeriod
PhD, Management Science and EngineeringShanghai University of Finance and Economics, SCAISept 2022 – Jul 2026
MS, Computer Software and TheoryJinan UniversitySept 2018 – Jul 2021
BS, Economic StatisticsDongbei University of Finance and EconomicsSept 2014 – Jul 2018

Research Experience

  • Visiting Student, Southern University of Science and Technology (SUSTech), Apr 2023 – Apr 2026, advised by Prof. Guanhua Chen

Publications

Auto-rendered from site.publications collection (existing template loop, already in cv.md).

Area 3: _publications/

Files to delete

FileReason
2024-11-12-spt.mdarXiv:2406.05946 belongs to Xiangyu Qi (not Zeguan Xiao); paper is not in resume
2025-06-07-jailbreak-bench.mdTitle fabricated; arXiv:2506.05038 is a preprint led by co-author Yutao Hou, not accepted

Files to fix

FileChanges
2021-11-07-bert4gcn.mdFix authors (Zeguan Xiao, Jiarun Wu, Qingliang Chen, Congjian Deng); fix URL to 2021.emnlp-main.724
2022-07-10-pruning-adaptperfusion.mdFull rewrite: correct title “Pruning Adatperfusion with Lottery Ticket Hypothesis”, correct authors (Jiarun Wu, Qingliang Chen, Zeguan Xiao, Yuliang Gu, Mengsi Sun), correct URL 2022.findings-naacl.123
2024-11-12-distract-jailbreak.mdFix URL from 2024.emnlp-main.1042 to 2024.emnlp-main.908
2025-04-29-seqar.mdFix title to “SeqAR: Jailbreak LLMs with Sequential Auto-Generated Characters”; fix authors (Yan Yang, Zeguan Xiao, Xin Lu, Hongru Wang, Xuetao Wei, Hailiang Huang, Guanhua Chen, Yun Chen); fix URL to 2025.naacl-long.42
2025-04-30-ar-checker.mdRewrite as “Modeling LLM Unlearning as an Asymmetric Two-Task Learning Problem” (ACL 2026, arXiv:2604.14808); authors: Zeguan Xiao, Shuai Li, Yiran Wang, Xuetao Wei, Jie Yang, Yun Chen, Guanhua Chen
2025-06-14-scalable-oversight.mdFix title to “Towards Bridging the Reward-Generation Gap in Direct Alignment Algorithms” (ACL Findings 2026, arXiv:2506.09457); authors: Zeguan Xiao, Yun Chen, Guanhua Chen, Kai Tang

Files to add

New filenamePaperVenueURL
2026-03-14-uncertainty-estimation.mdEnhancing Uncertainty Estimation in LLMs with Expectation of Aggregated Internal BeliefAAAI 2026arXiv:2509.01564, AAAI OJS: 40698
2026-04-19-representation-guided-unlearning.mdRepresentation-Guided Parameter-Efficient LLM UnlearningACL Findings 2026GitHub: reglu-open

Final publication list (8 papers)

#TitleVenueYear
1BERT4GCNEMNLP2021
2Pruning Adatperfusion with Lottery Ticket HypothesisNAACL Findings2022
3Distract Large Language Models for Automatic Jailbreak AttackEMNLP2024
4SeqAR: Jailbreak LLMs with Sequential Auto-Generated CharactersNAACL2025
5Enhancing Uncertainty Estimation in LLMs with Expectation of Aggregated Internal BeliefAAAI2026
6Modeling LLM Unlearning as an Asymmetric Two-Task Learning ProblemACL2026
7Representation-Guided Parameter-Efficient LLM UnlearningACL Findings2026
8Towards Bridging the Reward-Generation Gap in Direct Alignment AlgorithmsACL Findings2026

Non-Goals

  • No navigation or layout changes
  • No bilingual (/zh/) content changes
  • No new pages or sections
  • No style or theme changes

Verification Criteria

  • about.md renders with 肖泽管, School of Computing and Artificial Intelligence (SCAI), and includes LLM Unlearning
  • cv.md renders with SUFE PhD, Jinan University MS, DUFE BS
  • Publications page shows exactly 8 papers, all with correct titles and authors
  • No dead links to wrong arXiv IDs or wrong ACL Anthology entries