# Train Once, Forget Precisely: Anchored Optimization for Efficient Post-Hoc Unlearning

arXiv: https://arxiv.org/abs/2506.14515 · HTML: https://mlsanity.com/paper/2506.14515/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Machine unlearning, one of 12 directions in LLM safety and security.
- Builds on 9 of its 13 references that are in the map.
- First posted on arXiv 17 Jun 2025.

## Placed in

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

## Builds on (most cited first, 9 of 9)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Machine Unlearning](https://mlsanity.com/paper/1912.03817/) (2019)
- [Certified Data Removal from Machine Learning Models](https://mlsanity.com/paper/1911.03030/) (2019)
- [Remember What You Want to Forget: Algorithms for Machine Unlearning](https://mlsanity.com/paper/2103.03279/) (2021)
- [Forget-Me-Not: Learning to Forget in Text-to-Image Diffusion Models](https://mlsanity.com/paper/2303.17591/) (2023)
- [Learn to Forget: Machine Unlearning via Neuron Masking](https://mlsanity.com/paper/2003.10933/) (2020)
- [The Limits of Differential Privacy (and its Misuse in Data Release and Machine Learning)](https://mlsanity.com/paper/2011.02352/) (2020)
- [Recommendation Unlearning via Influence Function](https://mlsanity.com/paper/2307.02147/) (2023)
- [FAST: Feature Aware Similarity Thresholding for Weak Unlearning in Black-Box Generative Models](https://mlsanity.com/paper/2312.14895/) (2023)

## Built on by (most cited first, 2 of 2)

- [POUR: A Provably Optimal Method for Unlearning Representations via Neural Collapse](https://mlsanity.com/paper/2511.19339/) (2025)
- [Damnatio Memoriae: Adversarially and Selectively Forgetting Identities in the Embedding Space of Face Recognition Models](https://mlsanity.com/paper/2609.27115/) (2026)

## Abstract (arXiv)

As machine learning systems increasingly rely on data subject to privacy regulation, selectively unlearning specific information from trained models has become essential. In image classification, this involves removing the influence of particular training samples, semantic classes, or visual styles without full retraining. We introduce Forget-Aligned Model Reconstruction (FAMR), a theoretically grounded and computationally efficient framework for post-hoc unlearning in deep image classifiers. FAMR frames forgetting as a constrained optimization problem that minimizes a uniform-prediction loss on the forget set while anchoring model parameters to their original values via an ℓ2 penalty. A theoretical analysis links FAMR's solution to influence-function-based retraining approximations, with bounds on parameter and output deviation. Empirical results on class forgetting tasks using CIFAR-10 and ImageNet-100 demonstrate FAMR's effectiveness, with strong performance retention and minimal computational overhead. The framework generalizes naturally to concept and style erasure, offering a scalable and certifiable route to efficient post-hoc forgetting in vision models.