ResMem：学习你能学习的，并记忆其余部分
ResMem: Learn what you can and memorize the rest
AI总结：
本文提出ResMem算法，通过$k$近邻回归器拟合现有模型的残差以实现显式记忆，从而提升模型在视觉和自然语言处理基准上的泛化能力，并在理论上证明其测试风险优于基础预测器。
摘要：
现代神经网络出色的泛化性能部分归因于其隐式记忆复杂训练模式的能力。受此启发，我们探索了一种通过显式记忆来提升模型泛化的新机制。具体而言，我们提出了残差记忆（ResMem）算法，该方法通过使用基于$k$近邻的回归器拟合现有预测模型（如神经网络）的残差来增强该模型。最终预测即为原始模型与拟合残差回归器之和。从构造上看，ResMem能显式记忆训练标签。实验表明，ResMem在多个标准视觉和自然语言处理基准上，持续提升了原始预测模型的测试集泛化能力。理论上，我们构建了一个风格化的线性回归问题，并严格证明ResMem相比基础预测器能获得更优的测试风险。
英文摘要：
The impressive generalization performance of modern neural networks is attributed in part to their ability to implicitly memorize complex training patterns. Inspired by this, we explore a novel mechanism to improve model generalization via explicit memorization. Specifically, we propose the residual-memorization (ResMem) algorithm, a new method that augments an existing prediction model (e.g. a neural network) by fitting the model's residuals with a $k$-nearest neighbor based regressor. The final prediction is then the sum of the original model and the fitted residual regressor. By construction, ResMem can explicitly memorize the training labels. Empirically, we show that ResMem consistently improves the test set generalization of the original prediction model across various standard vision and natural language processing benchmarks. Theoretically, we formulate a stylized linear regression problem and rigorously show that ResMem results in a more favorable test risk over the base predictor.
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。
