Home Models Journal About GitHub
DOMAIN

Neuro-AI

Mapping biological neural systems to deep learning architectures. Understanding the brain to build better AI.

"To replicate the extraordinary capabilities of the human brain, we must first understand how it works."
MODELS
🧠
Brain Tumor MRI Classifier
4-class CNN · 96.2% accuracy · TensorFlow · Azure
Try demo
🔬
Virtual Lesion Analyzer
CNN vs ViT · Grad-CAM · Attention Rollout · SmoothGrad
Try demo ↓
📄
Medical Literature Q&A
RAG pipeline · PubMed · LangChain · Gemini 2.5
Try demo

LIVE DEMO

Virtual Lesion Analyzer

Open in new tab →
⏳ First load: The models (~420MB combined) may take 1–2 minutes to load on Hugging Face's free infrastructure. This is a research demo running on donated compute. Subsequent analyses will be faster.

RESEARCH TRACK
PHASE 1
Bio-Electric Foundation

Hodgkin–Huxley, LIF neuron models, visual stream mapping against CNN feature hierarchies.

PHASE 2
Computational Principles

Sparse coding, STDP learning rules, Hopfield networks, place cell representations.

PHASE 3
Cognition & Synthesis

Language processing, mechanistic interpretability, BCI neural signal decoder capstone.