In Japan, 2025, researchers at the RIKEN Center for Brain Science (CBS) have made a breakthrough in understanding how the brain simplifies complex sensory information. Taro Toyoizumi and Kensuke Yoshida developed a biologically realistic model, inspired by the fruit fly brain, that could revolutionize how we approach artificial intelligence and data processing.
The human brain is constantly bombarded with information from our senses. To avoid overload, it simplifies this data by reducing its dimensionality. This new model mimics this process, using a t-distributed stochastic neighbor embedding (t-SNE) algorithm adapted to reflect biological neural networks.
The model, consisting of three layers mirroring neuron groups in the fly brain, successfully replicated how fruit flies process smells. It also incorporates dopamine-dependent Hebbian plasticity, a key element in how the brain strengthens connections between neurons. This discovery could lead to more efficient and versatile AI systems capable of handling complex data with greater ease.
"The original t-SNE isn't biologically plausible - it's an engineering method rather than a neural network," explained Toyoizumi. "We rewrote the algorithm so that it mimicked a biological neural network." The team is now looking to apply their model to more complex systems, hinting at exciting possibilities for the future of AI and neuroscience.