2026 Neuro-AI Grand Hackathon

Team Structure and Research Topics

This hackathon brings together **5 labs** in a collaborative union with a total of **10 teams** participating. For a successful hackathon, each team must complete the required preparation items listed in the table below before the kick-off meeting. Refer to the document **'What is a Hackathon?'** section "Why is Preparation Critical?" for details.

Hackathon Teams and Research Topics

The following table summarizes the tentative research topics and objectives of the 10 teams identified so far. Team members and detailed topics will be finalized before the kick-off meeting.

Emotion Contextualized Perception

Research Topic: Verification of Emotion Contextualized Perception and Transformer model development

Expected Outcome: Benchmarking scores and trained model weights

Swift v3 Development

Research Topic: Swift v3 with longer context window and context-aware learning application

Expected Outcome: Swift v3 core logic and code implementation

fMRI VQ-VAE Training

Research Topic: fMRI VQ-VAE training and auto-regressive pre-training

Expected Outcome: Trained VQ-VAE model and generation examples

Affect-Contextualized Cognition

Research Topic: Affect-contextualized cognition and anxiety dynamics analysis (using CEBRA)

Expected Outcome: Mixed emotion classification accuracy and CEBRA visualization

Pretrained ECoG Model

Research Topic: Pretrained ECoG model: EEG Data Unification (Opensource + Closed-source)

Expected Outcome: ECoG model prototype based on integrated dataset

4D Brain Transformer

Research Topic: ABCD dataset-based 4D Spatiotemporal Self-Supervised Brain Transformer

Expected Outcome: Hybrid Parallelism optimized Transformer code

Benchmarking Study Design

Research Topic: New benchmarking study design and data acquisition (decoding verification)

Expected Outcome: Detailed research plan and pilot dataset

GPU Programming Optimization

Research Topic: Large-scale data processing pipeline construction through GPU programming optimization

Expected Outcome: Efficient data pipeline in HPC environment

Genetic Transformer

Research Topic: Neurogenetic feature analysis using Genetic Transformer

Expected Outcome: Genetic dataset-based Transformer model

To Be Determined

Research Topic: TBD

Expected Outcome: TBD

Essential Pre-Hackathon Preparation Checklist

All teams must complete the following items by January 15, 2026 and report to the team leader:

✅ Research Topic Definition

Testable hypothesis established
1-page proposal completed (including objectives, model structure)

✅ Data Curation and Preprocessing (Most Important)

Access permissions obtained for datasets to be used (e.g., ABCD Dataset)
Raw Data preprocessing completed to Model-Ready format
Backed up to external hard drive or cloud

✅ Environment and Role Assignment

GPU server access and library environment tested
Team roles assigned (core coding, data pipeline, result analysis)
Plan B (alternative scenario) discussed

The main hackathon event will be held at Wellpark Hotel, Gochang, Jeollabuk-do for 5 nights and 6 days from January 18-23, 2026.

Schedule	Content	Related Event
Preparation Deadline	Complete 3 essential preparations	Team Leader Confirmation
Kick-off Meeting	Finalize objectives and milestones	Calendar event
Main Hackathon Event	January 18-23, 2026	Wellpark Hotel, Gochang