# Assessing Per-Sample Membership Inference Vulnerability without Retraining

arXiv: https://arxiv.org/abs/2602.15919 · HTML: https://mlsanity.com/paper/2602.15919/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Membership inference attacks, one of 12 directions in LLM safety and security.
- Builds on 24 of its 35 references that are in the map.
- First posted on arXiv 17 Feb 2026.

## Placed in

- LLM safety and security / [Membership inference attacks](https://mlsanity.com/safety/membership-inference-attacks/)

## Builds on (most cited first, 24 of 24)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Deep Learning with Differential Privacy](https://mlsanity.com/paper/1607.00133/) (2016)
- [Membership Inference Attacks against Machine Learning Models](https://mlsanity.com/paper/1610.05820/) (2016)
- [Understanding Black-box Predictions via Influence Functions](https://mlsanity.com/paper/1703.04730/) (2017)
- [Understanding deep learning requires rethinking generalization](https://mlsanity.com/paper/1611.03530/) (2016)
- [Membership Inference Attacks From First Principles](https://mlsanity.com/paper/2112.03570/) (2021)
- [The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks](https://mlsanity.com/paper/1802.08232/) (2018)
- [Certified Data Removal from Machine Learning Models](https://mlsanity.com/paper/1911.03030/) (2019)
- [ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models](https://mlsanity.com/paper/1806.01246/) (2018)
- [What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation](https://mlsanity.com/paper/2008.03703/) (2020)
- [Certified Robustness to Adversarial Examples with Differential Privacy](https://mlsanity.com/paper/1802.03471/) (2018)
- [Does Learning Require Memorization? A Short Tale about a Long Tail](https://mlsanity.com/paper/1906.05271/) (2019)
- [Remember What You Want to Forget: Algorithms for Machine Unlearning](https://mlsanity.com/paper/2103.03279/) (2021)
- [Systematic Evaluation of Privacy Risks of Machine Learning Models](https://mlsanity.com/paper/2003.10595/) (2020)
- [Influence Functions in Deep Learning Are Fragile](https://mlsanity.com/paper/2006.14651/) (2020)
- [Auditing Differentially Private Machine Learning: How Private is Private SGD?](https://mlsanity.com/paper/2006.07709/) (2020)
- [Adversary Instantiation: Lower Bounds for Differentially Private Machine Learning](https://mlsanity.com/paper/2101.04535/) (2021)
- [The Privacy Onion Effect: Memorization is Relative](https://mlsanity.com/paper/2206.10469/) (2022)
- [Low-Cost High-Power Membership Inference Attacks](https://mlsanity.com/paper/2312.03262/) (2023)
- [Heterogeneous Differential Privacy](https://mlsanity.com/paper/1504.06998/) (2015)
- [Membership Inference Attack Using Self Influence Functions](https://mlsanity.com/paper/2205.13680/) (2022)
- [Parameters or Privacy: A Provable Tradeoff Between Overparameterization and Membership Inference](https://mlsanity.com/paper/2202.01243/) (2022)
- [Do Parameters Reveal More than Loss for Membership Inference?](https://mlsanity.com/paper/2406.11544/) (2024)

## Abstract (arXiv)

Recent work in the privacy literature shows that sample-targeted membership inference attacks (MIAs) significantly outperform untargeted approaches by a wide margin. Motivated by this observation, we address the following question: can the privacy vulnerability of individual training points be assessed without training shadow models? We show that per-sample exposure to MIA is governed not only by a point's loss, but also by a data-dependent geometric measure. In the linear setting, we derive a closed-form decomposition of individual black-box MIA vulnerability into a population leverage score and a residual loss term, making explicit how sample-dependent geometry translates into privacy exposure. Since the final layer of most modern architectures is linear, we extend this framework to deep networks and propose a surrogate score operating on last-layer representations that requires only a single trained model and no shadow models. Empirical evaluations across diverse datasets and architectures show that our score outperforms loss and gradient-norm baselines at identifying the highest-risk points under state-of-the-art attacks, providing a computationally efficient and theoretically grounded tool for per-sample privacy risk assessment.