Human learning progresses in stages. Humans benefit from training curricula. I will discuss the reasons for these phenomena, using neural networks trained with gradient descent as a model system. I will show that humans and neural network learning often unfolds in strikingly similar ways, following a similar path, and with similar sensitivity to structured curricula. I will discuss...
The human brain functions as a highly efficient communication network operating across multiple scales. Neurons exchange information locally through synapses, while distributed brain systems are connected by long-range pathways. Together, these connections form the connectome, the structural foundation for brain communication. Network neuroscience applies graph theory to reveal how patterns of...
