English(EN) Theory of Continual Learning Against Data Poisoning Attacks

新理论分析持续学习中的AI数据投毒

作者 PulseAugur 编辑部 · [1 个来源] · 2026-06-30 04:00

已开发出新的理论框架来分析持续学习（CL）中的数据投毒攻击和防御。研究人员将对手和防御者之间的交互构建为在线零和博弈，并确定了一个性能极限，即如果对手投毒了线性比例的任务，防御就会失败。该研究还探讨了不频繁攻击或有界噪声的情况，前者提出了一种任务到任务的验证机制，后者提出了一种鲁棒的防御机制，以最小化对投毒特征的敏感性。 AI

影响为理解和减轻持续学习系统中的数据投毒提供了理论基础，这对于LLMs和图像识别至关重要。

排序理由学术论文，详细介绍了分析持续学习中数据投毒的新理论框架。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 学术论文，详细介绍了分析持续学习中数据投毒的新理论框架。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 100 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Yiting Hu, Lingjie Duan · 2026-06-30 04:00

对抗数据投毒攻击的持续学习理论

arXiv:2606.29841v1 Announce Type: new Abstract: Continual learning (CL), where a model is trained on a sequence of data tasks, is increasingly being adopted across key fields such as large language models and image recognition, yet it remains highly vulnerable to data poisoning t…

本文如何被排名

报道来源 [1]

对抗数据投毒攻击的持续学习理论

相关实体

相关话题