PulseAugur
实时 11:13:25
English(EN) The BatchNorm Illusion: Diagnosing Normalization Artifacts in Machine Unlearning Evaluation

机器学习遗忘评估受 BatchNorm 伪影影响而存在缺陷

研究人员发现机器学习遗忘评估中存在一个显著的伪影,尤其影响使用 BatchNorm(批量归一化)的模型。这种被称为“BatchNorm 幻觉”的伪影,可以在不改变任何权重的情况下,通过在单次前向传播中改变模型的归一化状态,来逆转明显的遗忘指标。研究表明,这种幻觉可以将遗忘准确率提高多达 78 个百分点,并且可以通过使用 GroupNorm 代替 BatchNorm 来缓解。研究结果表明,由于这种测量偏差,先前的评估可能高估了遗忘方法的有效性。 AI

影响 突出了机器学习遗忘评估中的一个关键缺陷,可能使先前的结果无效,并需要新的评估协议。

排序理由 学术论文,详细介绍了机器学习遗忘评估方法中的一种新伪影。[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 →

机器学习遗忘评估受 BatchNorm 伪影影响而存在缺陷

本文如何被排名

Signal score
9 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
学术论文,详细介绍了机器学习遗忘评估方法中的一种新伪影。[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
Same-day
Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准

报道来源 [1]

  1. arXiv cs.LG TIER_1 English(EN) · Aaryaman Kalani, Murari Mandal, Dhruv Kumar, Mohan Kankanhalli, Yash Sinha ·

    BatchNorm 幻觉:机器学习模型遗忘评估中的归一化伪影诊断

    arXiv:2609.08901v1 Announce Type: new Abstract: Approximate machine unlearning aims to remove the influence of specific training data from a trained model without retraining from scratch. We identify a previously undocumented confound in how unlearning is evaluated on BatchNorm-b…