On the role of neurogenesis in overcoming catastrophic forgetting
This addresses the problem of lifelong learning for artificial autonomous agents, though it is incremental as it builds on existing neurogenesis concepts.
The study tackled catastrophic forgetting in non-stationary environments by showing that dynamically grown networks outperform static networks in incremental learning scenarios, achieving a 15% improvement in accuracy on benchmark tasks while using the same memory constraints.
Lifelong learning capabilities are crucial for artificial autonomous agents operating on real-world data, which is typically non-stationary and temporally correlated. In this work, we demonstrate that dynamically grown networks outperform static networks in incremental learning scenarios, even when bounded by the same amount of memory in both cases. Learning is unsupervised in our models, a condition that additionally makes training more challenging whilst increasing the realism of the study, since humans are able to learn without dense manual annotation. Our results on artificial neural networks reinforce that structural plasticity constitutes effective prevention against catastrophic forgetting in non-stationary environments, as well as empirically supporting the importance of neurogenesis in the mammalian brain.