About Me

I am a first-year Master's student majoring in Computer Science @ Tianjin University (TJU), advised by Prof. Bo Wang. Before that, I obtained my B.S. in Software Engineering @ Guangdong University of Foreign Studies (GDUFS), under the supervision of Prof. Xinying Qiu. Currently, I am visiting in iNLP Lab @ Singapore University of Technology and Design (SUTD), supervised by Prof. Wenxuan Zhang. I also have the pleasure of working with Shaoyang Xu and Yanlin Fu.
My research focuses on natural language processing, large language models, and trustworthy AI, with the goal of making AI alignment with both human values and real-world environments measurable, interpretable, and externalizable.

  • Measurable — “What gets measured gets improved.” [Alignment but Flattening]
  • Interpretable — “Understanding is the first step toward improvement.” [Self-debiasing Neurons]
  • Externalizable — “Actions speak louder than words.” [Work in Progress]
I am actively seeking industry internship opportunities. If you think my background could be a good fit, feel free to reach out at audbut0702@163.com.

🔥News

  • 2026.08: 🎉 Two papers (Alignment but Flattening & Beyond Alignment) were accepted to EMNLP 2026 Findings.
  • 2026.07: ✈️ Started my visiting at iNLP Lab @ SUTD.
  • 2026.01: 🎉 One paper (Let’s Simplify Step by Step) was accepted to EACL 2026 Findings.
  • 2025.09: 🎓 Started my master's journey at TJU.
  • 2024.09: 🚀 Joined Lionrock AI Lab as a Research Intern, focusing on world models.
  • 2024.09: 🎉 One paper (LCWL) was accepted to EMNLP 2024 Main Conference.
  • 2024.08: 🎉 One paper was accepted to IALP 2024.
  • 2024.07: 🎉 One paper (RISS) was accepted to CCL 2024.
  • 2024.06: 🏆 We won 1st place in Track 3 of the CCL Chinese Essay Fluency Evaluation (CEFE) Task 7.

Publications

* denotes equal contribution; † denotes the corresponding author.

Alignment but Flattening: Analyzing the Trade-off between Cultural Alignment and Diversity in LLMs

Jingshen Zhang, Shaoyang Xu, and Wenxuan Zhang†

EMNLP 2026 Findings. [Project][PDF]

Beyond Alignment: Value Diversity as a Collective Property in Multicultural Agent Systems

Shaoyang Xu, Jingshen Zhang, Long P. Hoang, Jinyuan Li, and Wenxuan Zhang†

EMNLP 2026 Findings. [PDF]

Modeling Implicit Conflict Monitoring Mechanisms against Stereotypes in LLMs

Jingshen Zhang, Bo Wang†, Yanlin Fu, Dongming Zhao, Ruifang He, Yuexian Hou, and Zifei Yu

arXiv 2026. [PDF]

Understanding the Self-Reflection Mechanisms of LLMs through Biased Attitude Associations

Jingshen Zhang, Bo Wang†, Boci Yang, Dongming Zhao, Ruifang He, Yuexian Hou, and Zifei Yu

arXiv 2026. [PDF]

Let’s Simplify Step by Step: Guiding LLM Towards Multilingual Unsupervised Proficiency-Controlled Sentence Simplification

Jingshen Zhang, Xinying Qiu†, Lifang Lu, Zhuhua Huang, Yutao Hu, Yuechang Wu, JunYu Lu

EACL2026 Findings. [PDF] [Code]

DualReward: A Dynamic Reinforcement Learning Framework for Cloze Tests Distractor Generation

Tianyou Huang*, Xinglu Chen*, Jingshen Zhang, Xinying Qiu†, and Ruiying Niu

CCL2025 Main. [PDF] [Code]

Label Confidence Weighted Learning for Target-level Sentence Simplification

Xinying Qiu† and Jingshen Zhang

EMNLP2024 Main. [PDF] [Code] [Poster]

Readability-guided Idiom-aware Sentence Simplification (RISS) for Chinese

Jingshen Zhang, Xinglu Chen, Xinying Qiu†, Zhimin Wang and Wenhe Feng

CCL2024 Main. [PDF] [Code] [Poster]

Controlling Cloze-test Question Item Difficulty with PLM-based Surrogate Models for IRT Assessment

Jingshen Zhang*, Jiajun Xie*, and Xinying Qiu†

Arxiv 2024. [PDF]

System Report for CCL24-Eval Task 7: Multi-Error Modeling and Fluency-Targeted Pre-training for Chinese Essay Evaluation

Jingshen Zhang*, Xiangyu Yang*, Xinkai Su, Xinglu Chen, Tianyou Huang, and Xinying Qiu†

CCL2024 Workshop. [PDF] [Code]

Cross-Lingual Word Alignment for ASEAN Languages with Contrastive Learning

Jingshen Zhang, Xinying Qiu†, Teng Shen, Wenyu Wang, Kailin Zhang, and Wenhe Feng

IALP2024. [PDF]

Internship

  • Singapore University of Technology and Design (SUTD)
    Visiting Student, Mar. 2025 - Now.
    Research Focus: Multicultural/Multilingual Large Language Models, Mechanistic Interpretability
  • Institute of Automation, Chinese Academy of Sciences (CASIA)
    Research Intern, Dec. 2025 - June. 2026, supervised by Prof. Guoqi Li.
    Research Focus: Vision-Language Model, Mechanistic Interpretability
  • Lionrock AI Lab, China Merchants Research Institute of Advanced Technology
    Research Intern, Sept. 2024 - Dec. 2024, supervised by Prof. Jiaxing Zhang and mentored by JunYu Lu.
    Research Focus: World Model

Podcast & Blogs & Life

  • Podcast: I host my own podcast, Jason’s Odyssey — Conversations on Frontier Technologies and Humanistic Thought.
  • Blogs: I consistently share my thoughts on trending topics and little moments from everyday life.
    • Paper Discussion: Stealing Reasoning Traces from Proprietary LLM APIs [Slides]
    • More moments from my daily life on [RedNote] (for fun and memories).
  • Life: I enjoy staying active, spending time in nature, and getting to know the local culture and customs wherever I go.
    • China Athletics Association (CAA) Level 1 recreational runner: Personal bests: 1:34 in the half marathon (4:30 min/km) and 3:49 in the marathon.
    • Solo, self-supported cycling tour around Qinghai Lake: I wrote a travelogue about the journey, which may be helpful if you are considering taking on the route yourself.