English(EN)Interference and Retention in Continual Learning
AI研究重新定义持续学习，超越记忆关注适应性
作者PulseAugur 编辑部·[11 个来源]·
近期研究论文探讨了AI模型中持续学习的复杂性，超越了简单的上下文管理，以解决模型能力随着世界变化而产生的根本性提升。研究调查了模型如何适应新领域和漂移数据，一些方法在快速适应方面表现出色，但在未来任务上表现下降，而另一些方法则更稳定地积累知识，但难以处理过时事实。一个突出的关键挑战是，当前的持续学习方法倾向于隐式地假设未来数据知识，而不是真正地无关紧要，这导致需要新的方法来平衡保留和适应性。
AI
arXiv:2604.14336v2 Announce Type: replace Abstract: Synaptic plasticity is metabolically expensive, yet animals continuously update their internal models without exhausting energy reserves. However, when artificial neural networks are trained, the network parameters are typically…
arXiv:2607.09202v1 Announce Type: cross Abstract: Continual learning commonly relies on post-hoc mechanisms such as replay, elastic regularization, or distillation. This work argues that forgetting should instead be modeled directly as interference between tasks. In the frozen-fe…
Continual learning commonly relies on post-hoc mechanisms such as replay, elastic regularization, or distillation. This work argues that forgetting should instead be modeled directly as interference between tasks. In the frozen-feature regime, forgetting from learning a new task …
arXiv cs.LG
TIER_1English(EN)·Rapha\"el Bayle, Martial Mermillod, Robert M. French·
arXiv:2511.08226v2 Announce Type: replace Abstract: In order to achieve Continual Learning (CL), the problem of catastrophic forgetting, one that has plagued neural networks since their inception, must be overcome. The evaluation of continual learning methods relies on splitting …
arXiv:2607.07847v1 Announce Type: new Abstract: As large language models (LLMs) become increasingly capable, the next question is how can we enable models to continually learn? Today, the field largely frames this as a problem of context management and mitigating forgetting. We a…
arXiv:2607.05609v1 Announce Type: new Abstract: The Continual Learning (CL) literature has long been driven by the goal of mitigating catastrophic forgetting. This objective rests on a pervasive, often unstated assumption: that a lifelong learner should approximate the Joint-Task…
arXiv:2507.08922v3 Announce Type: replace Abstract: Continual learning is an online paradigm where a learner continually accumulates knowledge from different tasks encountered over sequential time steps. Importantly, the learner is required to extend and update its knowledge with…
arXiv:2607.04711v1 Announce Type: new Abstract: Continual learning aims to progressively learn from a sequence of tasks, each containing a disjoint subset of classes, while preserving previously learned knowledge. Prompt-based continual learning methods propose to learn a small s…
The Continual Learning (CL) literature has long been driven by the goal of mitigating catastrophic forgetting. This objective rests on a pervasive, often unstated assumption: that a lifelong learner should approximate the Joint-Task Learning (JTL) solution and retain all previous…
Continual learning aims to progressively learn from a sequence of tasks, each containing a disjoint subset of classes, while preserving previously learned knowledge. Prompt-based continual learning methods propose to learn a small set of parameters, i.e., prompts, by associating …
Continual Learning bei LLMs: Catastrophic Forgetting erklärt Neues Wissen einbetten, ohne Altes zu verlieren? Catastrophic Forgetting ist kein Architekturfehler, sondern ein Gradienten-Risiko. Der Artikel erklärt die Mechanismen und warum Testpflicht vor Update Pflicht ist. https…
