A study on the plasticity of neural networks
One aim shared by multiple settings, such as continual learning or transfer learning, is to leverage previously acquired knowledge to converge faster on the current task. Usually this is done through fine-tuning, where an implicit assumption is that the network maintains its plasticity, meaning that the performance it can reach on any given task is not affected negatively by previously seen tasks. It has been observed recently that a pretrained model on data from the same distribution as the one it is fine-tuned on might not reach the same generalisation as a freshly initialised one. We build and extend this observation, providing a hypothesis for the mechanics behind it. We discuss the implication of losing plasticity for continual learning which heavily relies on optimising pretrained models.
Code (0)
등록된 구현이 없습니다.
Tasks
Continual LearningTransfer LearningSimilar Papers 제목 키워드 기반
Paired associative stimulation demonstrates alterations in motor cortical synaptic plasticity in patients with hepatic encephalopathy
Objective: Hepatic encephalopathy (HE) is a potentially reversible brain dysfunction caused by liver failure. Altered synaptic plasticity is supposed to play a major role in the pathophysiology of HE. Here, we used paire…
H-Mem: Harnessing synaptic plasticity with Hebbian Memory Networks
The ability to base current computations on memories from the past is critical for many cognitive tasks such as story understanding. Hebbian-type synaptic plasticity is believed to underlie the retention of memories over…
Question AnsweringEnvironmental variability and network structure determine the optimal plasticity mechanisms in embodied agents
The evolutionary balance between innate and learned behaviors is highly intricate, and different organisms have found different solutions to this problem. We hypothesize that the emergence and exact form of learning beha…
FormPLASTIC: Improving Input and Label Plasticity for Sample Efficient Reinforcement Learning
In Reinforcement Learning (RL), enhancing sample efficiency is crucial, particularly in scenarios when data acquisition is costly and risky. In principle, off-policy RL algorithms can improve sample efficiency by allowin…
reinforcement-learningReinforcement Learning (RL)An objective function for self-limiting neural plasticity rules
Self-organization provides a framework for the study of systems in which complex patterns emerge from simple rules, without the guidance of external agents or fine tuning of parameters. Within this framework, one can for…