# Space Over Statistic, Retraining Over Ranking: Auditing Training Labels Under Label-Only Backdoors

arXiv: https://arxiv.org/abs/2610.13051 · HTML: https://mlsanity.com/paper/2610.13051/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 56 of its 74 references that are in the map.
- First posted on arXiv 9 Oct 2026.

## Builds on (most cited first, 30 of 56)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain](https://mlsanity.com/paper/1708.06733/) (2017)
- [Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels](https://mlsanity.com/paper/1804.06872/) (2018)
- [Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning](https://mlsanity.com/paper/1712.05526/) (2017)
- [Distillation as a Defense to Adversarial Perturbations against Deep Neural Networks](https://mlsanity.com/paper/1511.04508/) (2015)
- [Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks](https://mlsanity.com/paper/1805.12185/) (2018)
- [Spectral Signatures in Backdoor Attacks](https://mlsanity.com/paper/1811.00636/) (2018)
- [STRIP: A Defence Against Trojan Attacks on Deep Neural Networks](https://mlsanity.com/paper/1902.06531/) (2019)
- [Detecting Backdoor Attacks on Deep Neural Networks by Activation Clustering](https://mlsanity.com/paper/1811.03728/) (2018)
- [Dataset Distillation by Matching Training Trajectories](https://mlsanity.com/paper/2203.11932/) (2022)
- [Backdoor Learning: A Survey](https://mlsanity.com/paper/2007.08745/) (2020)
- [Confident Learning: Estimating Uncertainty in Dataset Labels](https://mlsanity.com/paper/1911.00068/) (2019)
- [Hidden Trigger Backdoor Attacks](https://mlsanity.com/paper/1910.00033/) (2019)
- [Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks](https://mlsanity.com/paper/2007.02343/) (2020)
- [Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks](https://mlsanity.com/paper/2101.05930/) (2021)
- [Certified Defenses for Data Poisoning Attacks](https://mlsanity.com/paper/1706.03691/) (2017)
- [Label-Consistent Backdoor Attacks](https://mlsanity.com/paper/1912.02771/) (2019)
- [Poisoning Web-Scale Training Datasets is Practical](https://mlsanity.com/paper/2302.10149/) (2023)
- [Identifying Mislabeled Data using the Area Under the Margin Ranking](https://mlsanity.com/paper/2001.10528/) (2020)
- [A new Backdoor Attack in CNNs by training set corruption without label poisoning](https://mlsanity.com/paper/1902.11237/) (2019)
- [Dataset Distillation with Infinitely Wide Convolutional Networks](https://mlsanity.com/paper/2107.13034/) (2021)
- [Blind Backdoors in Deep Learning Models](https://mlsanity.com/paper/2005.03823/) (2020)
- [Witches' Brew: Industrial Scale Data Poisoning via Gradient Matching](https://mlsanity.com/paper/2009.02276/) (2020)
- [Backdoor Defense via Decoupling the Training Process](https://mlsanity.com/paper/2202.03423/) (2022)
- [SentiNet: Detecting Localized Universal Attacks Against Deep Learning Systems](https://mlsanity.com/paper/1812.00292/) (2018)
- [Universal Litmus Patterns: Revealing Backdoor Attacks in CNNs](https://mlsanity.com/paper/1906.10842/) (2019)

## Abstract (arXiv)

Some backdoor attacks never touch an image. FLIP only changes labels: on CIFAR-10, flipping 3% of them makes a ResNet-32 send 99.7% of triggered source-class images to the attacker's class. The usual defenses look for the poison inside a model trained on those same labels. So when they struggle, is the problem the test they run, or the space they run it in? We answer with a controlled swap: we keep the test, the labels and the metric fixed, change only the feature space, and judge each defense by retraining the victim on what it keeps. The space decides. On dog→cat the same nearest-neighbor check scores AUROC 0.546 on the poisoned model's features, 0.967 on frozen DINOv2 features that never saw a label, and 0.995 on the same network trained on clean labels, an oracle control rather than a defense. We reproduce SPECTRE, the defense FLIP's authors recommend, to 90.94 / 0.97 against its published 91.47 / 1.7, then price it: on truck→deer it deletes 20% of every class, while a prototype audit on the frozen features deletes 6% for higher clean accuracy and lower attack success. On three harder pairs that audit leaves about a quarter of the attack standing. A count read off encoder geometry and clean labels, before any attack is run, separates the pair where the audit removes the attack from those where it only partly does, though on a single zero-count pair. Three results warn against the easy story. A better detection score does not mean a safer model: on two of four pairs Confident Learning scores the higher AUROC, and on dog→cat leaves 2.6 times the attack success. The prototype family, our best static performer, is the first an adaptive attacker slips past. And past a point, catching more poisons stops reducing attack success. Check labels in a space that never learned them, and judge a defense by the retrained model, not by its detection score.