Zekun Wu · Research systems 5 projects · 2024–2026 PhD · Saarland University
Zekun Wu

PhD student
Computational Interaction Lab, Saarland Univ.

03
— Systems

Projects

Five systems and studies, 2024 – 2026. Each with paper, code, or demo where available.

01

System · 2026

Trace2Flow: From Review to Reuse

Turning a completed AI-agent execution into an editable workflow graph users can inspect, validate, and reuse.

Under construction
Paper Demo Code
Trace2Flow: an agent's execution trace is translated into an editable workflow graph for inspection and reuse
Fig. 01 — Trace to workflow graph

02

System · 2026

Gaze-based Child Reading Assistant

An adaptive AI reading companion using real-time eye-tracking.

Under construction
Paper Demo
Gaze-based child reading assistant interface
Fig. 02 — Reading companion

03

Study · HAXD 2026

AI Prompt Design for Influencing User Beliefs

How AI response detail and confidence shape human judgment.

Paper Code
Survey interface for the belief-change study
Fig. 03 — Study interface

04

System · ETRA 2025

RelEYEance: Real-Time Gaze-Based AI Reliance Detection

Detecting and calibrating user reliance on AI through eye-tracking.

Paper Code
RelEYEance teaser figure
Fig. 04 — Reliance pipeline

05

Model · IEEE Trans. Human-Machine Systems

Highlight-Informed Saliency Model (HISM)

Predicting how visual highlights guide user attention in interfaces.

Paper Code
HISM model architecture
Fig. 05 — Model architecture