Haltere

Github Repository: Here↗

So, a new project huh. This one is really special one. Kind of complicated, but a really new and fun idea.

In September 2026, Google and Janelia released a full wiring map of a male fruit fly's brain and nerve cords. It has a LOT of cells (~166,000 neurons and ~125 million connections). This is a connectome. But, as in normal neural networks, the map is not enough. You also need the weights, or how much a neuron effects the output. You can only find these by simulating the connectome and comparing output until you get similar outputs. Haltere is essentially a tool for this, we give it the map, it will give us tools to simulate it and change weights on the connectome.

Check out the Initial Plan at the time of this Intro Log: Here↗

Okay, so first is the Research Stage of this Project. I have already done some preliminary research, but doing deeper research is essential before starting this project.

Since I study at The University of Hong Kong, I have access to plenty of Good University Professors that specialize in Neural Biology. I have already talked to two, and they told me the idea is good, but needs further research before starting.

I have found some research papers that I can read before getting started. Here is my list:

I will get started by reading one per day. Will update after sufficient Research.

Comments