arXiv Artificial Intelligence

One-Point Contraction: Erasing Representational Separability toward Irreversible Deep Forgetting

1 hour ago 05.08.2026 07:00 👁 1 views
One-Point Contraction: Erasing Representational Separability toward Irreversible Deep Forgetting

Quick summary

arXiv:2507.07754v3 Announce Type: replace-cross Abstract: Machine unlearning is usually evaluated by what the classifier outputs: forget-set accuracy, confidence, membership-inference scores. We show that this is not enough. Across 14 representative unlearning methods on CIFAR-10 and SVHN, a single linear map fitted on a held-out calibration set, with no access to the forgotten data, reverses the unlearning in seconds and recovers forget-set accuracy to within a few percent of the original model. Recovered features even support pixel-level reconstruction through a generic decoder. We call this

Key takeaways

  • arXiv:2507.07754v3 Announce Type: replace-cross Abstract: Machine unlearning is usually evaluated by what the classifier outputs: forget-set accuracy, confidence, membership-inference scores.
  • Across 14 representative unlearning methods on CIFAR-10 and SVHN, a single linear map fitted on a held-out calibration set, with no access to the forgotten data, reverses the unlearning in seconds and recovers forget-set accuracy to within a few percent of the original model.
  • Recovered features even support pixel-level reconstruction through a generic decoder.

Why it matters

“One-Point Contraction: Erasing Representational Separability toward Irreversible Deep Forgetting” should be evaluated beyond branding and benchmark scores. Its practical importance will emerge in task accuracy, latency, unit cost, safety and integration with real workflows.

Kaynak sitede devamını oku: arXiv Artificial Intelligence ↗