Life sciences · Structure → signals → behavior
Connectomics
How can the same parts produce different behaviors? Change the connections.
A connectome is a map of connections between neurons. Connectomics asks how that network is organized—and how its structure helps explain what an animal does.
1. Learn to read the map
Neurons are nodes
A neuron receives signals and sends signals to other cells. A network diagram represents each neuron as a node.
Synapses are links
A chemical synapse passes a signal from a sending cell to a receiving cell. Arrows show that direction. One pair of neurons can share many synapses.
Signals change activity
Excitatory input tends to increase a receiving neuron's activity; inhibitory input tends to decrease it. The effect also depends on the receiving cell and its state.
A real map: the 2024 FlyWire brain
Researchers reconstructed 139,255 neurons and 54.5 million chemical synapses in one adult female fruit fly brain. Electron microscopy reveals tiny structures; computer reconstruction and human proofreading trace the cells and their connections.
A wiring map is a foundation, not a recording of activity. To explain behavior, researchers also need measurements of signals, connection strengths, body mechanics, and the environment. Read the original research (Nature, 2024).
A controlled thought experiment
2. Rewire a two-sensor robot
This robot is a model inspired by Valentino Braitenberg's thought experiments. It has two light sensors and two wheels. It is not a simulation of a fruit fly brain. A brighter sensor either speeds up or slows down a connected motor.
The steering rule: a faster left wheel turns the robot right; a faster right wheel turns it left. Trace the wires before you press Run.
Move the light: click or drag in the arena, or focus the arena and use arrow keys. The dotted line records the robot's path. At an edge the robot reappears opposite.
Choose the connections
+ increases speed · − decreases speed
- Left sensor
- 0%
- Right sensor
- 0%
- Left motor
- 0
- Right motor
- 0
Sensors: relative light intensity. Motors: model units/s. Trials use the same start; switching wiring pauses and resets the robot.
3. Predict → test → explain
Reset the trial. The light starts ahead and to the robot's left. Choose B · Crossed +. Which way should the robot initially turn?
Make a prediction, then step the simulation and compare the motor readings.
- Compare A and B. Keep the starting position and light fixed. Which single change reverses the initial turn?
- Compare B and C. Both turn toward the light, but what happens to their speed as they approach?
- Move the light equally far from both sensors. Explain why equal wheel speeds produce a straight path.
Reveal the explanation and model limits
With the light to the left, the left sensor receives a stronger signal. In B it drives the right motor faster, turning the robot left. A sends it to the left motor, turning right. In C, stronger left input slows the left motor, also turning left—but reducing speed near the light.
This model uses a smooth distance-based light signal, instantaneous connections, and ideal wheels. It has no spikes, learning, noise, or real neural circuit. Labels such as “fear” and “love” describe a path that looks familiar; they do not show that a robot feels an emotion. Near the light, inhibitory motors can stop when both sensor signals reach their maximum.