# Stake the Points: Structure-Faithful Instance Unlearning

arXiv: https://arxiv.org/abs/2603.12915 · HTML: https://mlsanity.com/paper/2603.12915/

## 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 27 of its 42 references that are in the map.
- First posted on arXiv 13 Mar 2026.

## Placed in

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

## Builds on (most cited first, 27 of 27)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [A Survey on Bias and Fairness in Machine Learning](https://mlsanity.com/paper/1908.09635/) (2019)
- [Understanding Black-box Predictions via Influence Functions](https://mlsanity.com/paper/1703.04730/) (2017)
- [Machine Unlearning](https://mlsanity.com/paper/1912.03817/) (2019)
- [Deep Leakage from Gradients](https://mlsanity.com/paper/1906.08935/) (2019)
- [Adversarial Examples Are Not Bugs, They Are Features](https://mlsanity.com/paper/1905.02175/) (2019)
- [Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks](https://mlsanity.com/paper/1911.04933/) (2019)
- [SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation](https://mlsanity.com/paper/2310.12508/) (2023)
- [Visual Classification via Description from Large Language Models](https://mlsanity.com/paper/2210.07183/) (2022)
- [Unrolling SGD: Understanding Factors Influencing Machine Unlearning](https://mlsanity.com/paper/2109.13398/) (2021)
- [Fast Yet Effective Machine Unlearning](https://mlsanity.com/paper/2111.08947/) (2021)
- [Fast Machine Unlearning Without Retraining Through Selective Synaptic Dampening](https://mlsanity.com/paper/2308.07707/) (2023)
- [Boundary Unlearning](https://mlsanity.com/paper/2303.11570/) (2023)
- [Federated Unlearning via Class-Discriminative Pruning](https://mlsanity.com/paper/2110.11794/) (2021)
- [Towards Adversarial Evaluations for Inexact Machine Unlearning](https://mlsanity.com/paper/2201.06640/) (2022)
- [Deep Unlearning via Randomized Conditionally Independent Hessians](https://mlsanity.com/paper/2204.07655/) (2022)
- [Learning to Unlearn: Instance-wise Unlearning for Pre-trained Classifiers](https://mlsanity.com/paper/2301.11578/) (2023)
- [Language Guided Concept Bottleneck Models for Interpretable Continual Learning](https://mlsanity.com/paper/2503.23283/) (2025)
- [Decoupled Distillation to Erase: A General Unlearning Method for Any Class-centric Tasks](https://mlsanity.com/paper/2503.23751/) (2025)
- [Few-Shot Unlearning by Model Inversion](https://mlsanity.com/paper/2205.15567/) (2022)
- [LoTUS: Large-Scale Machine Unlearning with a Taste of Uncertainty](https://mlsanity.com/paper/2503.18314/) (2025)
- [RainbowPrompt: Diversity-Enhanced Prompt-Evolving for Continual Learning](https://mlsanity.com/paper/2507.22553/) (2025)
- [ESC: Erasing Space Concept for Knowledge Deletion](https://mlsanity.com/paper/2504.02199/) (2025)

## Abstract (arXiv)

Machine unlearning (MU) addresses privacy risks in pretrained models. The main goal of MU is to remove the influence of designated data while preserving the utility of retained knowledge. Achieving this goal requires preserving semantic relations among retained instances, which existing studies often overlook. We observe that without such preservation, models suffer from progressive structural collapse, undermining both the deletion-retention balance. In this work, we propose a novel structure-faithful framework that introduces stakes, i.e., semantic anchors that serve as reference points to maintain the knowledge structure. By leveraging these anchors, our framework captures and stabilizes the semantic organization of knowledge. Specifically, we instantiate the anchors from language-driven attribute descriptions encoded by a semantic encoder (e.g., CLIP). We enforce preservation of the knowledge structure via structure-aware alignment and regularization: the former aligns the organization of retained knowledge before and after unlearning around anchors, while the latter regulates updates to structure-critical parameters. Results from image classification, retrieval, and face recognition show average gains of 32.9%, 22.5%, and 19.3% in performance, balancing the deletion-retention trade-off and enhancing generalization.