Structurally Flexible Neural Networks: Evolving the Building Blocks for General Agents

arXiv:2404.15193v1 Announce Type: cross
Abstract: Artificial neural networks used for reinforcement learning are structurally rigid, meaning that each optimized parameter of the network is tied to its specific placement in the network structure. It also means that a network only works with pre-defined and fixed input- and output sizes. This is a consequence of having the number of optimized parameters being directly dependent on the structure of the network. Structural rigidity limits the ability to optimize parameters of policies across …

abstract agents artificial artificial neural networks arxiv building cs.ai cs.lg cs.ne general meaning network networks neural networks placement reinforcement reinforcement learning type