English · 繁體中文
I’m a PhD researcher in Civil and Environmental Engineering at Lehigh University. My work sits at the intersection of behavioral science, LLM-based simulation, and quantitative evaluation. A central part of my research is understanding and modeling human decision-making. I evaluate whether LLMs capture meaningful decision processes, not only whether their answers sound plausible, but whether their perceptions and constraints relate coherently to the choices they make. I also work with agent-based modeling, which is widely used to simulate how individual decisions accumulate over time and interact with larger social, financial, and physical systems. My current direction is to move from traditional rule-based agents toward LLM-driven agents that can respond to richer information, memory, and changing conditions. More broadly, I’m interested in how human behavior can be simulated in a way that remains empirically grounded while interacting with and changing the external world.
Verified by ai-profile · refreshed daily by GitHub Actions from public Git evidence · Unknown stays Unknown.
| FLOODABM | Agent-based model coupled with catastrophe flood modeling — household flood adaptation in the Passaic River Basin (NJ, 2011–2023) |
LLM Agents · AI Harness / Governed Agents · Agent-Based Modeling · Multi-Agent Systems · Catastrophe Modeling · Decision-Making Under Risk · Human-Environment Interaction
- Leveraging Large Language Models for Agent-Based Simulation of Human-Water System Interactions — Water Resources Research, 62(6), e2025WR042111 (2026)
| ai-research-skills | 5 plugins · 14 research skills · install in one command |
| agent-collab-skills | Multi-agent orchestration — task splitter, output reconciler, debate, shared memory, acceptance gate |
| zotero-skills | Programmatic Zotero CRUD — search, add, classify, annotate, organize literature |
| research-hub | AI-operable research workspace for Zotero, Obsidian, and NotebookLM — CLI, MCP, REST, dashboard |
| academic-writing-skills | Rigorous academic paper writing, revision, and submission — field-agnostic with per-paper journal overrides |
| codex-delegate | Single-agent delegation skill used by the marketplaces above |
| ai-profile | Local-first AI collaboration analytics — explicit Git provenance in, privacy-safe cards and an interactive provider dashboard out |
Selected merged pull requests:
| ZhuLinsen/daily_stock_analysis | Added Taiwan (台股) market support — market detection & routing, an institutional-flow (三大法人) data fetcher, and TWD currency / dividend fixes · merged PRs |
| langchain-ai/openwiki | Restricted ~/.openwiki permissions on Windows, where chmod is a no-op · #367 |
| BuilderIO/agent-native | Routed AGENT_ENGINE=codex-cli to the Codex runner instead of crashing · #1362 |
| Nanako0129/coralline | Fixed auto-layout over-wrapping — measure display columns, not bytes · #10 |
| awesome-agentic-ai-zh |  7-stage agentic-AI roadmap · trilingual (繁中 canonical / 简中 / EN) · 240+ curated resources, from LLM basics to multi-agent production |
Open to full-time Summer 2027 internships (approximately late May through mid-August) in LLM evaluation, agent systems, behavioral simulation, and AI for science. F-1 student; CPT eligible. Contact: wec324@lehigh.edu
perceive → reason → verify → act · repeat