AI RESEARCH
Is Backpropagation Optimal? When Synthetic Gradients Improve Sample Efficiency
arXiv CS.LG
•
ArXi:2605.27946v1 Announce Type: cross Backpropagation is the default learning rule for artificial neural networks and is often treated as the settled approach whenever differentiability is available. In this work, we revisit this convention through a theoretical lens of sample efficiency. We