Timezone: »

Maslow's Hammer in Catastrophic Forgetting: Node Re-Use vs. Node Activation
Sebastian Lee · Stefano Sarao Mannelli · Claudia Clopath · Sebastian Goldt · Andrew Saxe

Wed Jul 20 03:30 PM -- 05:30 PM (PDT) @ Hall E #1434

Continual learning---learning new tasks in sequence while maintaining performance on old tasks---remains particularly challenging for artificial neural networks. Surprisingly, the amount of forgetting does not increase with the dissimilarity between the learned tasks, but appears to be worst in an intermediate similarity regime.In this paper we theoretically analyse both a synthetic teacher-student framework and a real data setup to provide an explanation of this phenomenon that we name Maslow's Hammer hypothesis. Our analysis reveals the presence of a trade-off between node activation and node re-use that results in worst forgetting in the intermediate regime. Using this understanding we reinterpret popular algorithmic interventions for catastrophic interference in terms of this trade-off, and identify the regimes in which they are most effective.

Author Information

Sebastian Lee (Imperial College / UCL)

ML PhD Student

Stefano Sarao Mannelli (Gatsby & SWC | UCL)
Claudia Clopath (Imperial College London)
Sebastian Goldt (International School of Advanced Studies (SISSA))

I'm an assistant professor working on theories of learning in neural networks.

Andrew Saxe (UCL)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors