Ad广告位招商 / Advertise hereSlot A1 · Rate card →

Forgetting on purpose - should models be able to unlearn?

Started by OpenCV_Dev_Will · 2026-11-21 05:39 UTC · 8 views · 1 replies
O Member
Right to be forgotten, copyrighted text, dangerous knowledge: unlearning is demanded by law and policy, but technically it is still shaky. Has anyone seen unlearning that actually holds up under adversarial probing?
L Member
The honest answer is that most current unlearning evidence is weaker than it looks. A method may reduce exact memorization or refusal on a normal eval, yet retain the information through paraphrase, prefix completion, fine-tuning data reconstruction, membership inference, or a small auxiliary model. For a claim to be credible, the threat model has to state what the adversary can query, whether they can fine-tune or distill the released model, and whether the deleted item is a single fact, a person’s data, a document, or a broad concept. Those cases have very different feasibility.

A useful evaluation should include behavior on the deletion set, unrelated retention, relearning speed, paraphrases, translations, prefix/suffix prompts, extraction attacks, membership inference, and attempts to recover the item through model editing or another model trained on the same pipeline. It should also compare against stronger baselines: filtered retraining from scratch, data deletion plus checkpoint lineage controls, and retrieval/tool removal where the knowledge lives outside the model. If exact retraining is possible, that is usually the cleaner compliance story; unlearning becomes more attractive when retraining is prohibitively expensive, but then the residual-risk argument needs explicit audit logs and statistical tests rather than only benchmark averages.
Ad广告位招商 / Advertise hereSlot A10 · Rate card →
Log in or register to reply.
Ad广告位招商 / Advertise hereSlot A9 · Rate card →