面向自适应持续学习的神经拟态元可塑性
Neuromimetic metaplasticity for adaptive continual learning
AI总结：
针对深度神经网络的灾难性遗忘问题，提出受人类工作记忆启发的神经拟态元可塑性模型，通过混合不同灵活度的突触实现无预处理/后处理的持续学习，兼顾记忆容量与性能，且对数据投毒攻击具有鲁棒性。
摘要：
基于深度神经网络（DNN）模型的传统智能系统因灾难性遗忘问题，在实现类人持续学习方面面临挑战。本文提出一种受人类工作记忆启发的元可塑性模型，无需任何预处理或后处理，即可让DNN实现无灾难性遗忘的持续学习。该方法的核心要点是实现从稳定型到灵活型的不同类型突触，并将它们随机混合，以训练具有不同灵活度的突触连接。这一策略让网络即使在输入长度发生意外变化的情况下，也能成功学习连续的信息流。该模型无需额外训练或结构修改，就能在记忆容量和性能之间实现平衡权衡，动态分配记忆资源以同时保留新旧信息。此外，该模型利用赫布重复效应强化重要数据的留存，通过选择性过滤错误记忆，展现出对数据投毒攻击的鲁棒性。
英文摘要：
Conventional intelligent systems based on deep neural network (DNN) models encounter challenges in achieving human-like continual learning due to catastrophic forgetting. Here, we propose a metaplasticity model inspired by human working memory, enabling DNNs to perform catastrophic forgetting-free continual learning without any pre- or post-processing. A key aspect of our approach involves implementing distinct types of synapses from stable to flexible, and randomly intermixing them to train synaptic connections with different degrees of flexibility. This strategy allowed the network to successfully learn a continuous stream of information, even under unexpected changes in input length. The model achieved a balanced tradeoff between memory capacity and performance without requiring additional training or structural modifications, dynamically allocating memory resources to retain both old and new information. Furthermore, the model demonstrated robustness against data poisoning attacks by selectively filtering out erroneous memories, leveraging the Hebb repetition effect to reinforce the retention of significant data.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
