Xiaoyue Yu Information

Information

I’m Xiaoyue (Helen) Yu 禹晓月.

I’m a senior undergraduate student at the University of Pittsburgh, majoring in Information Science.
I also have a dual degree in Industrial Engineering from Sichuan University, China.

I care about design, Human-AI Collaboration, and Education.


CV.pdf

Open CV.pdf

PantherWays: Conversational AI Advisor

In development

PantherWays is a retrieval-augmented conversational academic advisor for University of Pittsburgh undergraduates. It elicits a student's goals, retrieves real Pitt data via RAG to ground its recommendations, and produces two consent-gated outputs: a student summary and an advisor brief delivered through a meeting-request flow.

My work includes interviewing and coordinating with academic advisors, psychology experts, English department faculty, and computer science faculty; designing the user interaction model; shaping the frontend UI; structuring the database content; and developing prompts for grounded, advisor-ready recommendations.

LLM Annotator

In development

LLM Annotator was built by the McLearn Lab in 2026 to leverage large language models (LLMs) for qualitative annotation of natural language data.

My work includes polishing the frontend UI, refactoring the code structure, and exploring interaction patterns that make the annotation workflow more approachable for non-technical users.

Language Learning: WhatsApp Chat Bot

In development

Language Learning: WhatsApp Chat Bot is a WhatsApp-based language learning chatbot designed to help children in Ecuador learn Kichwa, a local Indigenous language.

My work includes designing and building the MySQL database schema, developing game-based learning activities, and connecting the chatbot experience with the Twilio API.

The Anxiety Time Machine

Feb 6–8, 2026

A web platform that reframes today’s worries through the lens of a lifetime. Share an anxiety or a happy moment, then turn the temporal dial to read how people across different ages felt the same way — and revisit your own emotional history in a personal journey.

Built with Boyi Sun and Ruoyu Wu for She Innovates 2026. React · Flask · SQLite · Gemini API.

View on Devpost

The Playable Archive

Oct 2025 – Feb 2026

A research project that turns digital archives into social virtual museum worlds, making large cultural collections explorable, curatable, and conversational across platforms.

https://www.playablearchive.com/

Game Jam: Black Friday

The Jam @ GDC

Black Friday is a game we made for The Jam @ GDC, a 36-hour on-site game jam at Moscone Center that brought together 100 jammers from around the world. With the theme "Celebrate," the jam invited teams to create games as part of the Festival of Gaming.

The Jam @ GDC was hosted in the GDC Commons in the Festival Hall and powered by Tencent Institute of Games as the Sustenance Sponsor.

Play on itch.io

Game Jam: Paperworkx

Pitt's Games 4 Social Impact 2025

Paperworkx is a puzzle game about a robot on Earth going through the legal process to immigrate to Mars. Players fill out immigration paperwork and must answer enough questions correctly to pass.

We made Paperworkx for Pitt's Games 4 Social Impact 2025, an in-person University of Pittsburgh game jam focused on social impact. The 2025 jam theme was Migration, and Paperworkx ranked #1 in Narrative Design.

My work focused on Unity game development, core narrative design, and technical implementation.

Play on itch.io

Publications

  1. Under Review

    Assertive versus Defensive Automated Driving: Impacts on Prefrontal Activity and Intervention Tendency

    Danduo Huang, Wenjing Qin, Xiaoyue Yu, Yuzhou Wu, Yanchun Wang, Yaoyu Fu, Rong Yin, Hui Jiang, Dan Pu, Kang Li, Xiandi Wang, Xiaomei Tan.

    Under Review at Transportation Research Part F: Traffic Psychology and Behaviour. Preprint: SSRN.

    doi:10.2139/ssrn.7157788

  2. June 2025

    Blood Biomarker-Based Machine Learning Triaging Model for Predicting the Risk of Cognitive Decline Progression

    Jiuyue Yin, Xiaoyue Yu, Jiayi Pan, Guangwu Qian.

    2025 IEEE 2nd International Conference on Deep Learning and Computer Vision (DLCV).

    doi:10.1109/DLCV65218.2025.11088776

  3. May 2025

    Drivers’ Intervention Tendency and Trust Dynamics During Interactions with Automated Vehicles

    Yuzhou Wu, Danduo Huang, Xiaoyue Yu, Yanchun Wang, Wenjing Qin, Xiaomei Tan.

    HCI International 2025 Posters.

    doi:10.1007/978-3-031-94159-7_44

Lab Experiences

  1. McLearn Lab

    Human-Computer Interaction Institute (HCII), Carnegie Mellon University

    Research Intern

    PI: Prof. Bruce M. McLaren · Mentors: Daniel Pollock, Hayden Stec

    Apr 2026 – Present

    • Extended an LLM-powered tool that automates qualitative annotation of natural-language research data for HCI and learning-science workflows.
    • Refactored the Python codebase into modular components, redesigned the researcher-facing UI, and integrated the OpenAI API with checkpoint-based recovery for long-running tasks.
    • Ran an independent data-mining study on Decimal Point, an educational game, analyzing student logs with DataShop metrics (AIC, BIC, RMSE).
  2. Learning Technologies Lab

    School of Computing and Information, University of Pittsburgh

    Undergraduate Research Fellow & Assistant

    PI: Dr. Dmitriy Babichenko

    Aug 2025 – Present

    • Awarded the 2026 NOUR Fellowship to lead independent HCI research under faculty mentorship.
    • Leading a WhatsApp language-learning chatbot (Flask + Twilio) for Kichwa language preservation — designing the MySQL schema, E-R diagram, and game flow around Indigenous cultural values.
    • Led user interviews, IRB compliance, and thematic analysis for two VR projects, including The Playable Archive (Secondary Study Coordinator) and a VR locomotion study.
  3. Pervasive HCI Lab

    Tsinghua University, Beijing, China

    Research Intern (Remote)

    Mentor: Chang Liu

    Dec 2025

    • Researched eye-tracking integration with LLMs to infer user cognitive intent while reading AI-generated text.
    • Investigated RLHF-inspired alignment by modeling simulated versus empirical gaze patterns to capture implicit reading preferences.
    • Built a Python backend prototype linking gaze-stream data with LLM pipelines to generate real-time context annotations without explicit prompts.
  4. Human-Centered AI Lab (HAIL)

    Sichuan University–Pittsburgh Institute, Chengdu, China

    Research Assistant

    PI: Dr. Xiaomei Tan

    Oct 2023 – Jun 2025

    • Investigated human factors and user trust dynamics in Level 2 autonomous driving under hazard scenarios.
    • Managed experimental protocols, participant recruitment, and fNIRS cerebral hemodynamic data collection — contributing to a published HCII 2025 poster and an SSRN submission under review.