The stability-plasticity dilemmaIt is a constraint of the nervous system, which exists in both artificial and biological nervous systems and can be understood as a trade-off between stability and plasticity to achieve a state of balance.

Learning in parallel and distributed systems requires plasticity to integrate new knowledge, while stability is used to prevent forgetting previous knowledge; excessive plasticity will lead to the continuous forgetting of previously encoded data, while excessive stability will hinder the effective encoding of the data at the synaptic level.

In most cases, neural computing has now solved the problems of excessive plasticity and excessive stability.