Wen Yu Chang
Wen Yu Chang

Wen Yu (Morris) Chang

Ph.D. Student · Computer Science and Information Engineering
National Taiwan University

I am a Ph.D. student at NTU CSIE, advised by Prof. Yun-Nung (Vivian) Chen at the MiuLab. My research focuses on building conversational agents that remember, adapt, and act: agent memory systems, language model personalization, and conversational agents & dialogue systems. I joined the Ph.D. program directly from NTU's M.S. program in Data Science, after a B.B.A. in Accounting at NTU, and spent nearly three years building LLM-based agentic systems in industry.

  • Agent Memory
  • LLM Personalization
  • Dialogue Systems
  • Conversational Agents
Email Google Scholar GitHub LinkedIn CV (PDF)

News

Publications

Full list on Google Scholar.

2026

  1. Figure 1 preview
    EMNLP 2026 Findings

    When Users Don't Ask: Benchmarking Context-Driven Memory Retrieval in Conversational Agents

    Wen Yu Chang, Yun-Nung Chen

    Findings of the Association for Computational Linguistics: EMNLP 2026

    arXiv

  2. Figure 1 preview
    ACL 2026 Findings

    VisTW: Benchmarking Vision-Language Models for Taiwanese Mandarin in Taiwan

    Zhi Rui Tam, Yung-Yu Shih, Yen-Wei Lee, Ya-Ting Pai, Wen Yu Chang, Yun-Nung Chen

    Findings of the Association for Computational Linguistics: ACL 2026

    PaperarXiv

  3. Figure 1 preview
    IWSDS 2026

    FlowSwitch: A State-Aware Framework for Workflow Transitions in Adaptive Dialogue Agents

    Wen Yu Chang, Luning Qiu, Yi-Hung Liu, Yun-Nung Chen

    16th International Workshop on Spoken Dialogue System Technology (IWSDS 2026)

    Paper

  4. Figure 1 preview
    IWSDS 2026

    The Context Trap: Why End-to-End Audio Language Models Fail Multi-turn Dialogues

    Zhi Rui Tam, Wen Yu Chang, Yun-Nung Chen

    16th International Workshop on Spoken Dialogue System Technology (IWSDS 2026)

    PaperCode

2025

  1. Figure 3 preview: SalesAgent pipeline with occupation-based strategies
    ASRU 2025

    From Simulation to Strategy: Automating Personalized Interaction Planning for Conversational Agents

    Wen Yu Chang, Tzu-Hung Huang, Chih-Ho Chen, Yun-Nung Chen

    IEEE Automatic Speech Recognition and Understanding Workshop (ASRU 2025)

    arXivCode

  2. Figure preview
    IWSDS 2025

    Exploring Personality-Aware Interactions in Salesperson Dialogue Agents

    Sijia Cheng, Wen Yu Chang, Yun-Nung Chen

    15th International Workshop on Spoken Dialogue Systems Technology (IWSDS 2025)

    PaperarXiv

  3. Figure preview
    MMM 2025 Demo

    Using Language Models to Generate and Forget the Narrative Memories of an Assistive Robot

    Angel F. Garcia Contreras, Wen Yu Chang, Seiya Kawano, Yun-Nung Chen, Koichiro Yoshino

    31st International Conference on Multimedia Modeling (MMM), Demo Paper

    Paper

2024

  1. Figure 1 preview
    ACL 2024 Findings

    Injecting Salesperson's Dialogue Strategies in Large Language Models with Chain-of-Thought Reasoning

    Wen Yu Chang, Yun-Nung Chen

    Findings of the Association for Computational Linguistics: ACL 2024

    PaperarXiv

Experience

Wistron Corporation

Industry-Academia Research Collaborator

Dec. 2025 – PresentTaiwan
  • Helped build a human annotation pipeline for evaluating and improving LM-based systems.
  • Contributed to benchmark construction: task formulation, annotation protocol design, and data quality control.
  • Supported evaluation workflows that measure model behavior under realistic task settings.
  • Benchmarking
  • Human Annotation
  • LLM Evaluation

Taiwan Semiconductor Manufacturing Company (TSMC)

Industry-Academia Research Collaborator

Dec. 2024 – Feb. 2026Taiwan
  • Research and system design for knowledge graph construction, inference pipelines, and SOP-grounded retrieval.
  • Investigated retrieving, following, and reasoning over large-scale standard operating procedures in industrial settings.
  • Designed a knowledge-intensive pipeline connecting structured knowledge, retrieval modules, and downstream reasoning.
  • Knowledge Graphs
  • Retrieval
  • Industrial SOPs

MaverickP, Inc.

AI Engineer

June 2023 – Mar. 2026Taipei, Taiwan
  • Developed LLM-based agentic systems automating human workflows such as product market analysis and domain-specific QA.
  • Built product specification knowledge graphs to support factual, controllable generation.
  • Integrated databases, web search, APIs, and knowledge graphs to improve factual grounding of LLM agents.
  • Translated business requirements into executable AI workflows and system-level prototypes.
  • LLM Agents
  • Tool Use
  • Knowledge Graphs
  • Neo4j

RIKEN Guardian Robot Project

Research Intern · Knowledge Acquisition and Dialogue Research Team

July 2023 – Sept. 2023Kyoto, Japan
  • Agent Memory
  • Human–Robot Interaction
  • LLM Simulation

National Taiwan University

Teaching Assistant · Applied Deep Learning

Mar. 2022 – Dec. 2023Taipei, Taiwan
  • Designed and graded NLP assignments, including Chinese news summarization and neural text generation.
  • Mentored students on model implementation, experimental design, and debugging.

National Taiwan University

Teaching Assistant · Artificial Intelligence

Mar. 2023 – June 2023Taipei, Taiwan
  • Developed an assignment on representation learning covering PCA, autoencoders, and denoising autoencoders.

Education

Feb. 2025 – Present

Ph.D. in Computer Science and Information Engineering

National Taiwan University · Direct Ph.D. Program

Advisor: Prof. Yun-Nung (Vivian) Chen
Admitted directly from the M.S. program in Data Science

Mar. 2022 – Jan. 2025

M.S. Program in Data Science

National Taiwan University

Transferred to the Ph.D. program prior to degree conferral
Advisor: Prof. Yun-Nung (Vivian) Chen
GPA: 4.12 / 4.3

Sept. 2016 – Jan. 2022

B.B.A. in Accounting

National Taiwan University

GPA: 4.00 / 4.3

Skills

NLP / ML
PyTorch, LangChain, Neo4j, vLLM
Programming
Python, Bash, C++, C
Languages
Mandarin (native), English — TOEFL iBT 107/120, GRE 329/340