The community has recently developed various training-time defenses to counter neural backdoors introduced through data poisoning. In light of the observation that a model learns poisonous samples responsible for the backdoor easier than benign samples, these approaches either use a fixed threshold of the training loss for splitting or iteratively learn a reference model as an oracle for identifying benign samples.
arXiv:2607. 05748v1 Announce Type: new Abstract: The community has recently developed various training-time defenses to counter neural backdoors introduced through data poisoning.
By Qi Zhao, Christian Wressnegger
arXiv:2609.15029v1 Announce Type: cross
Abstract: Backdoor poisoning attacks add poisoned examples to otherwise-clean finetuning data, pairing a trigger with a target behavior that the model learns t...
By Aashiq Muhamed, Mona T. Diab, Virginia Smith, Andrew Ilyas, Matthew Jagielski
The paper evaluates two training‑time data poisoning attacks—label flipping and backdoor poisoning—on MNIST and Fashion‑MNIST using Logistic Regression, Linear SVM, and Random Forest classifiers. Label flipping degrades performance most for Logistic Regression and Linear SVM, while Random Forest remains relatively stable. Backdoor poisoning achieves near‑perfect attack success rates across all models while largely preserving clean‑test accuracy, highlighting the stealthy nature of targeted backdoors.
By Toshif Khan (Minot State University), Muhammad Abusaqer (Minot State University)
arXiv:2602. 04899v2 Announce Type: replace-cross Abstract: We present a data poisoning attack -- Phantom Transfer -- with the property that, even if you know precisely how the poison was placed into an otherwise benign dataset, you cannot filter it out.
By Andrew Draganov, Tolga H. Dur, Anandmayi Bhongade, Mary Phuong
Backdoor attacks compromise training data so that a model retains clean accuracy but predicts an attacker-chosen target on triggered inputs. At very low poisoning rates, only a few samples convey the trigger--target association, making poison-sample selection critical.
The paper demonstrates that a single poisoned data point can successfully create a backdoor in linear models and ReLU neural networks without needing detailed knowledge of the training data. It establishes provable conditions under which this one‑poison attack works with high probability, achieving zero backdooring error while leaving the model’s normal performance largely unaffected. The attack relies only on coarse geometric bounds of the input space and training parameters.
By Thorsten Peinemann, Paula Arnold, Sebastian Berndt, Thomas Eisenbarth, Esfandiar Mohammadi
SAGE is a defense against clean‑label data poisoning that relies on a very small set of verified examples—both clean and poisoned—rather than a large clean set. It trains a generic feature extractor on a separate dataset and then uses a non‑parametric, similarity‑weighted prediction to flag poisoned training examples. Experiments on standard benchmarks show that even a handful of verified poisoned examples give a substantial advantage, and that the distribution of verified clean examples across classes is more important than their sheer number.
By Chaeeun Han, Soodeh Atefi, Yevgeniy Vorobeychik, Aron Laszka
arXiv:2607. 26849v1 Announce Type: cross Abstract: As large language models (LLMs) are deployed in high-stakes domains, adversaries may poison training data to implant backdoors: hidden triggers that covertly manipulate model behavior at inference time.
By Anthony Hughes, Nicole Xing, Collin Francel, Andy Kim, Andrew Draganov
arXiv:2606. 04929v1 Announce Type: new Abstract: LLM post-training proceeds through multiple stages, e.
By Jack Sanderson, Yihan Wang, Xiaoqian Lu, Gautam Kamath, Yiwei Lu
As large language models (LLMs) are deployed in high-stakes domains, adversaries may poison training data to implant backdoors: hidden triggers that covertly manipulate model behavior at inference time. We ask whether a defender can recover such a trigger under realistic affordances, namely white-box access to the weights and knowledge of the behavior of concern, but no training data, no trusted reference model, no knowledge of the trigger, and no certainty that the model is poisoned.
arXiv:2606. 15123v1 Announce Type: cross Abstract: We study the task of CVE-conditioned exploit generation, where a model drafts proof-of-concept (PoC) exploits given software vulnerability context.
By Yiwei Chen, Lichi Li, Kai Cheung, Vinny Parla, Ganesh Sundaram