# The Nearest Target Is the Wrong One: Target Separation in Arc2Face Identity Unlearning

arXiv: https://arxiv.org/abs/2608.30087 · HTML: https://mlsanity.com/paper/2608.30087/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Machine unlearning, one of 12 directions in LLM safety and security.
- Builds on 20 of its 23 references that are in the map.
- First posted on arXiv 30 Aug 2026.

## Placed in

- LLM safety and security / [Machine unlearning](https://mlsanity.com/safety/machine-unlearning/)

## Builds on (most cited first, 20 of 20)

- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Deep Learning Face Attributes in the Wild](https://mlsanity.com/paper/1411.7766/) (2014)
- [Machine Unlearning](https://mlsanity.com/paper/1912.03817/) (2019)
- [Erasing Concepts from Diffusion Models](https://mlsanity.com/paper/2303.07345/) (2023)
- [InstantID: Zero-shot Identity-Preserving Generation in Seconds](https://mlsanity.com/paper/2401.07519/) (2024)
- [AdaFace: Quality Adaptive Margin for Face Recognition](https://mlsanity.com/paper/2204.00964/) (2022)
- [Unified Concept Editing in Diffusion Models](https://mlsanity.com/paper/2308.14761/) (2023)
- [PhotoMaker: Customizing Realistic Human Photos via Stacked ID Embedding](https://mlsanity.com/paper/2312.04461/) (2023)
- [Ablating Concepts in Text-to-Image Diffusion Models](https://mlsanity.com/paper/2303.13516/) (2023)
- [A Survey of Machine Unlearning](https://mlsanity.com/paper/2209.02299/) (2022)
- [Selective Amnesia: A Continual Learning Approach to Forgetting in Deep Generative Models](https://mlsanity.com/paper/2305.10120/) (2023)
- [Boundary Unlearning](https://mlsanity.com/paper/2303.11570/) (2023)
- [Anti-DreamBooth: Protecting users from personalized text-to-image synthesis](https://mlsanity.com/paper/2303.15433/) (2023)
- [SER-FIQ: Unsupervised Estimation of Face Image Quality Based on Stochastic Embedding Robustness](https://mlsanity.com/paper/2003.09373/) (2020)
- [Arc2Face: A Foundation Model for ID-Consistent Human Faces](https://mlsanity.com/paper/2403.11641/) (2024)
- [Circumventing Concept Erasure Methods For Text-to-Image Generative Models](https://mlsanity.com/paper/2308.01508/) (2023)
- [Verification of Machine Unlearning is Fragile](https://mlsanity.com/paper/2408.00929/) (2024)
- [GraFIQs: Face Image Quality Assessment Using Gradient Magnitudes](https://mlsanity.com/paper/2404.12203/) (2024)
- [SMI: Statistical Membership Inference for Reliable Unlearned Model Auditing](https://mlsanity.com/paper/2602.01150/) (2026)
- [PIU: Proximity-guided Identity Unlearning in ID-Conditioned Diffusion Models](https://mlsanity.com/paper/2605.22311/) (2026)

## Abstract (arXiv)

Unlearning an identity from a face-conditioned generator by redirecting its conditioning embedding can silently fail if the redirected output is still verified as the original person. We show that this failure depends on a controllable choice of how far the redirection target lies from the forget identity in recognition space, and that the most intuitive target, the nearest neighbour, is the one most likely to cause it. We audit Arc2Face with a locked ArcFace protocol and a projection adapter that redirects identity conditioning before generation. On a hard-neighbour stress test built from the hardest 0.5% of eligible identities, four target-selection policies show a monotonic response: clean forgetting rises from 9/30 groups under the nearest hard target to 30/30 under the least similar one. Mean forget-identity re-identification falls from 51.9 to 0.0 while mean retention stays flat. This reflects successful redirection rather than outputs becoming unverifiable: 710 of 720 least-sim-hard generations arrive at the chosen target, with no leakage to unrelated identities. Re-verifying identical images with an independent recogniser (AdaFace) preserves that trend, correlating at r=0.94, arguing against a verifier artefact. Target separation is thus a first-order, reportable design variable for identity unlearning.