arXiv cs.AI· Kathrin Korte, Christian Metzner, Akshay Bedhotiya, Viet Anh Khoa Tran, Eleni Nisioti, Sebastian Risi·· 11 小时前AI 评分18
持续学习中对任务的隐含假设
Implicit Assumptions About Tasks in Continual Learning
AI 导读
持续学习通常将数据集划分为任务并按顺序呈现，但数据相似度、变化速率和呈现顺序等因素究竟指代底层数据、数据划分还是学习器本身，往往未被明确。研究通过受控的教师—学生实验表明，以不同粒度划分同一底层轨迹、改变呈现顺序、调整数据相似度结构，都会导致不同的测量结果。作者还审查了代表性基准，指出在预训练语料未公开的基础模型中，底层数据的属性由模型初始状态继承。
正文 · 原文
Abstract:Continual learning is the study of how a model accumulates knowledge from data received over time, rather than all at once. In practice, this is typically achieved by dividing a dataset into tasks and presenting them sequentially. Explanations of the resulting measurements take into account factors such as the similarity between segments, the rate at which the data changes and the order of presentation. Each of these factors can be defined in terms of the underlying data, the partition imposed on it, or the learner that encounters it. These levels do not necessarily coincide, and it is usually left implicit which level a reported quantity refers to. We separate these and identify which properties acquire a value at each level. In controlled teacher--student experiments, we demonstrate that partitioning the same underlying trajectory at different levels of granularity, changing the order of presentation, and varying the similarity structure of the underlying data all result in different measured outcomes. We review representative benchmarks to determine which of these properties are stated. Where the underlying data cannot be inspected, as in foundation models built on undisclosed pre-training corpora, its properties are inherited by the model's initial state, and the training applied on top of it can be described.
| Comments: | Accepted to the Workshop on Continual Learning in the Era of Foundation Models and Embodied Agents at NeurIPS 2026 |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.41792 [cs.LG] |
| (or arXiv:2610.41792v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.41792 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Kathrin Korte [view email]
[v1]
Fri, 30 Oct 2026 14:22:26 UTC (2,753 KB)
来源：arXiv cs.AI · arxiv.org