情景式观察：用强化学习偏置感知提升少样本与持续学习并防遗忘
Episodic looking - using reinforcement learning to bias perception enhances few-shot and continual learning with protection from forgetting
AI 导读
arXiv 论文提出基于互补学习系统（CLS）架构的情景式观察方法：由短期记忆在长期记忆产生的鲁棒泛化感知空间中工作，长期记忆负责统计泛化，短期记忆专注记忆近期输入并避免干扰。
正文 · 原文
Abstract:There is a fundamental contradiction between the desire for models that can generalize, and models that can continuously adapt to recent input with changing statistics. These conditions, captured in few-shot and continual learning settings, can bias or disrupt learned representations. This harms generalization to a broader context, leaving the model unable to overcome input variance.
This paper proposes a solution based on the Complementary Learning System architecture (CLS), in which a Short-Term Memory (STM) delegates generalization to a slow, statistical Long-Term Memory (LTM). Delegation is achieved by working in a space of robust, generalized perceptions produced by the LTM. With this separation, the STM can focus on memorizing recent input and avoiding interference between memories.
We show that on an image classification task, the resulting CLS model is more resistant to forgetting in a domain-incremental continual learning setting than a baseline ResNet LTM, and generalizes more effectively in a few-shot setting than other continual learners.
In the continual setting, overall accuracy is slightly worse than existing continual learning models, but with a different mix of performance characteristics. Results analysis suggests that the model is limited by the representations and representational plasticity of the LTM. In future work, we will seek to address this by replaying STM memories to the LTM.
| Comments: | 21 pages, 8 figures |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| MSC classes: | 68T05 (Primary) 68T07, 68Q32 (Secondary) |
| ACM classes: | I.2.6; I.2.6; I.5.1 |
| Cite as: | arXiv:2610.38577 [cs.LG] |
| (or arXiv:2610.38577v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.38577 arXiv-issued DOI via DataCite |
Submission history
From: David Rawlinson [view email]
[v1]
Wed, 28 Oct 2026 10:09:11 UTC (1,824 KB)
来源：arXiv cs.AI · arxiv.org