发表机构
Zhejiang University; Ant Group（浙江大学; 蚂蚁集团）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文研究课程学习中的灾难性遗忘，理论证明其源于任务最优分歧而不可完全消除，提出IV-EWC方法，通过影响函数构建验证集并动态正则化，在数学推理任务上平均减少162%遗忘。
AI 中文摘要
课程学习已在众多领域得到广泛应用。然而，其有效性本质上受到灾难性遗忘的限制，这种遗忘是由课程任务之间模型参数分布的偏移所驱动的。在本文中，我们基于课程学习已确立的有效性，研究这种训练范式中的灾难性遗忘现象。我们对参数更新动态的理论分析表明，课程学习中的灾难性遗忘源于任务最优点的分歧，而这种分歧通常对于课程学习更快的收敛是必要的；因此，遗忘无法被完全消除。基于这一发现，我们增强训练过程并提出IV-EWC，它将弹性权重巩固（EWC）纳入课程学习目标，以抑制数学推理（一个典型的课程学习场景）中的灾难性遗忘。IV-EWC使用影响函数从课程的训练数据中构建一个有代表性的验证集，用于在训练过程中驱动动态正则化。我们进一步提出扩展的理论分析，表明基于EWC的正则化方法能缓解课程学习中的灾难性遗忘，从而为IV-EWC提供理论支持。在三个骨干模型和三个基准上的实证评估表明，课程学习表现出灾难性遗忘。IV-EWC缓解了这一问题，相对于普通课程学习平均减少了162%的遗忘，并产生了正向的向后迁移，表现为在具有挑战性的任务上训练后，在较简单任务上的性能有所提升。
英文摘要
Curriculum learning has found broad application across numerous domains. Nevertheless, its effectiveness is intrinsically curtailed by catastrophic forgetting, driven by the shifts in model parameter distributions between curriculum tasks. In this paper, we investigate the phenomenon of catastrophic forgetting in this training paradigm, building on the established efficacy of curriculum learning. Our theoretical analyses of parameter update dynamics demonstrate that catastrophic forgetting in curriculum learning stems from the divergence of task optima, which is generally essential to the faster convergence of curriculum learning; therefore, forgetting cannot be completely eliminated. Based on this finding, we augment the training process and propose IV-EWC, which incorporates Elastic Weight Consolidation (EWC) into the curriculum learning objective to curb catastrophic forgetting in mathematical reasoning, a prototypical curriculum learning scenario. IV-EWC employs the influence function to construct a representative validation set from the curriculum's training data, which is used to drive dynamic regularization during training. We further present an extended theoretical analysis to show that EWC-based regularization methods mitigate catastrophic forgetting in curriculum learning, thereby providing theoretical support for IV-EWC. Empirical evaluations on three backbone models and three benchmarks indicate that curriculum learning exhibits catastrophic forgetting. IV-EWC alleviates this issue, reducing forgetting by 162% on average relative to vanilla curriculum learning and yielding positive backward transfer, as evidenced by improved performance on easier tasks after subsequent training on challenging tasks.
