Overview
The Vertically Integrated Project Program (VIPP) functions as applied research rather than traditional coursework. Students should embrace three core principles: "Learn by building. Start from the data and the task, and learn the theory and tools as you go," maintain transparency by sharing progress within the lab community, and prioritize comprehensive documentation.
Weekly Rhythm & Deliverables
Students meet Monday afternoons in the lab. By 12:00 PM Mondays, submit a one-page PDF progress brief to both the VIPP Slack channel and your project channel.
The brief should contain five sections:
- Done: Concrete work completed, results, screenshots, links
- Learned: New concepts and tools acquired; surprising discoveries
- Blocked: Obstacles encountered and attempted solutions
- Next: Clear, testable steps for the coming week
- Links: GitHub commits, papers, datasets, notebooks, demos
Post finalized weekly to-dos on Slack after the meeting. Throughout the week:
- Execute your plan and share notable discoveries on Slack
- Monitor Slack each morning and respond within 24 hours when tagged
Time Expectations
VIPP requires minimum 3 hours per credit per week of focused project work, preferably completed in-lab for supervisor feedback. Budget additional time for background learning in Python, machine learning, or music theory to maintain productive project hours. The philosophy emphasizes learning through iteration: start with end goals, prototype small solutions, and deepen understanding iteratively.
Responsible AI Tool Usage
Language models accelerate learning when used conscientiously. Students must understand everything submitted.
- Probe and verify: Question outputs, request mathematical explanations, demand source citations, ask for beginner-friendly explanations
- Validate: Test code, write minimal tests, sanity-check results
- Explain thoroughly: Demonstrate code comprehension and decision-making to supervisors
- Credit assistance: Note AI tools in your 1-pager; share effective workflows with peers; acknowledge uncertainty
Tools & Onboarding Checklist
Essential tools include: Python (or project-relevant language), Jupyter/VS Code, Git & GitHub, relevant DAWs/music software, and Slack with notifications enabled.
Onboarding steps:
- Join Slack and enable VIPP channel notifications
- Configure Git/GitHub; create private repos with clear README.md documentation
- Verify Python functionality, notebook capability, and dependency installation
- Gather domain-specific materials (papers, datasets, theoretical resources)
- Push initial prototype, however minimal (data load, visualization, stub class, basic evaluation)
Documentation Standards
Follow the "three D's" approach:
- Daily: Maintain lightweight notes on attempted solutions and command syntax
- Dev: Commit frequently with meaningful messages; include "How to run" instructions in README.md
- Digest: Weekly PDFs serve as digests, keep them visual and link-rich
Quality documentation appears "short, skimmable, and reproducible" with clear inputs/outputs, version information, and links to papers, repos, datasets, and commits. When sharing links, provide context explaining relevance and beneficiaries.
Working Style: Engineering Mindset
- Begin with minimal scope; validate tiny functional pieces immediately
- Test early by running repositories and replicating figures on toy datasets
- Iterate: build, test, reflect, refine
- Document visibly through screenshots, video clips, and plotted metrics
Collaboration & Culture
The lab embraces interdisciplinary work with musicians, artists, dancers, and peers across fields. Communication should be generous, ask concise questions, demonstrate attempts, maintain Slack responsiveness. Practice good citizenship through kind code review, appropriate credit attribution, source citation, and license/privacy respect. "Curiosity and play are the core part of the craft."
Deliverables & Milestones
- Weekly: One-page PDF progress brief, Slack to-dos, active repository
- Mid-term: Brief demonstration of simplified progress toward semester goals
- Final: Working prototype and reproducible write-up or presentation
- Optional: Lightning talks during lab showcases for significant findings
Evaluation emphasizes consistency, documentation clarity, collaboration quality, appropriate technical depth, and a functional demonstration.
Getting Started: First Tasks
- Confirm your assigned project thread
- Draft a one-paragraph goal for the next two weeks (problem statement, specific target, success metrics)
- Develop a minimal prototype this week; report findings in your first Monday 1-pager