LoTUS让模型忘记特定数据，无需重新训练，适合大规模实际应用。
LoTUS: Large-Scale Machine Unlearning with a Taste of Uncertainty
- 通过信息论边界平滑预测概率，缓解数据记忆带来的过自信问题。
- 在五个数据集上优于8种基线方法，尤其在ImageNet1k上表现突出。
- 提出新评估指标RF-JSD，支持无重训条件下的真实场景测试。
我们提出LoTUS，一种新型机器遗忘（Machine Unlearning, MU）方法，可在不从头训练的前提下消除预训练模型中特定训练样本的影响。该方法将模型预测概率平滑至信息论上限，缓解因数据记忆导致的过自信问题。我们在Transformer和ResNet18模型上，对八种基线方法在五个公开数据集上进行了评估。除了常规MU基准测试外，还在ImageNet1k这一大规模数据集上进行了评估，该场景下重训不切实际，更贴近真实应用。为此，我们引入了新的无重训詹森-香农散度（Retrain-Free Jensen-Shannon Divergence, RF-JSD）指标，以支持真实环境下的评估。实验结果表明，LoTUS在效率与效果上均优于现有最先进方法。代码已开源：https://github.com/cspartalis/LoTUS。
原文摘要 · Abstract (English)
We present LoTUS, a novel Machine Unlearning (MU) method that eliminates the influence of training samples from pre-trained models, avoiding retraining from scratch. LoTUS smooths the prediction probabilities of the model up to an information-theoretic bound, mitigating its over-confidence stemming from data memorization. We evaluate LoTUS on Transformer and ResNet18 models against eight baselines across five public datasets. Beyond established MU benchmarks, we evaluate unlearning on ImageNet1k, a large-scale dataset, where retraining is impractical, simulating real-world conditions. Moreover, we introduce the novel Retrain-Free Jensen-Shannon Divergence (RF-JSD) metric to enable evaluation under real-world conditions. The experimental results show that LoTUS outperforms state-of-the-art methods in terms of both efficiency and effectiveness. Code: https://github.com/cspartalis/LoTUS.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。