AxGS Lab: Student & Collaboration Inquiries

AxGS Lab focuses on AI x Games Systems, Trustworthy AI, and Optimization. We run projects from problem definition to implementation, evaluation, and documentation. We look for people who want to finish strong and ship real outcomes.

Undergraduate RA / Capstone / Collaboration Weekly project meetings Code, experiment logs, and docs together
Student Inquiry Collaboration Proposal Publications

AxGS Lab Direction

How We Work

Implementation First

AxGS Lab prioritizes building real systems, services, and games through vibe coding. We do not stop at concepts or demos; projects are pushed toward working, end-to-end deliverables.

Vibe Coding + Systems Mindset

Speed matters, but so do data quality, inference cost, deployment stability, and user experience. Rapid implementation is paired with production-oriented system thinking.

Fast Feedback

Weekly reviews identify bottlenecks early, convert blockers into concrete action items, and keep implementation velocity high.

Tracks

Who We Work With

Undergraduate RA

8-12 week mini projects. Basic Python, Git, and data-handling experience is recommended.

Capstone Integration

Capstone topics can be connected to lab tracks with stronger technical design and evaluation.

Graduate Preparation

Research reading, reproduction experiments, and structured reporting for M.S./Ph.D. readiness.

Industry Collaboration

From practical problem reformulation to KPI-driven PoC and deployment strategy.

Academic Collaboration

Co-authoring, benchmark sharing, and experiment partitioning with clear contribution boundaries.

Data/Infra Partnership

Data access, evaluation environments, and log-based system analysis collaborations are welcome.

Execution Process

Flow

  • Step 1
    Problem setup: align target metrics, scope, and expected deliverables.
  • Step 2
    Technical design: break down data/model/evaluation/deployment tasks.
  • Step 3
    Implementation and validation: weekly iteration with explicit review checkpoints.
  • Step 4
    Final packaging: clean repo, technical report/slides, and publication potential assessment.

Student Guide

Before You Reach Out

You are a good fit if

  • You prefer building working things over discussing ideas indefinitely
  • You can leave visible progress every week through commits and runnable outputs
  • You surface blockers early and respond well to direct technical feedback
  • You want at least one concrete outcome: portfolio project, capstone result, service, or game prototype

Minimum baseline

  • Basic Python and the ability to install and run libraries on your own
  • Basic Git usage: clone, commit, push
  • Willingness to debug using error messages, documentation, and LLM tools
  • Ability to read README files, API docs, and implementation examples

What the first 4 weeks look like

  • Week 1: topic selection, environment setup, repo structure
  • Week 2: one minimal feature implemented with a runnable output
  • Week 3: refine data/prompt/evaluation flow and fix bottlenecks
  • Week 4: bring the project to a demo-ready state with README and slide draft

Expected deliverables

  • One runnable repository
  • A README that explains setup, scope, and core functionality
  • Screenshots, a short demo video, or a live link
  • When appropriate: slides, technical notes, report, or paper draft

How to Apply

Inquiry Format

Student Inquiry

Use subject format: [Course][StudentID] Inquiry Topic. Review is much faster when you include your current level and at least one existing artifact.

  • Student ID / year / course context
  • Interest track(s): LLM, Trustworthy AI, Optimization
  • Tooling background: Python, SQL, Git, etc.
  • Weekly availability and goal for this semester
  • At least one link: GitHub, portfolio, prototype, or write-up
  • Target outcome: project, portfolio, research track
Send Student Inquiry

Collaboration Proposal

Use subject format: [Collaboration Proposal] Affiliation_Topic.

  • Problem statement and current baseline
  • Data and environment availability
  • Expected outcome and timeline
  • Collaboration type: co-authoring, PoC, system build
Send Proposal

FAQ

Common Questions

How demanding is it week to week?

Even a mini project requires steady weekly time. This is not a passive participation model; you are expected to build, document, and leave visible progress every week.

Can I still apply if my coding is weak?

Yes, but basic Python, Git, and debugging effort are required. You do not need to be polished already, but you do need to ship weekly progress and respond to feedback.

Can I apply if I mainly want to build games?

Yes. But games here are treated as full implementation projects, including system structure, AI behavior, data flow, evaluation, and user-facing delivery.

Is using LLMs alone enough?

No. Prompting is only one part. We also care about data, evaluation, cost, logs, user experience, and deployment stability. The goal is a working system, not a prompt demo.

Is publication mandatory?

No. Portfolio projects, capstone outputs, prototypes, and service demos are all valid goals. But every track is expected to leave concrete artifacts such as a repo, README, and demo.

How fast do you reply?

Typically within 2 business days. During exam periods, response may be delayed.