arXiv Machine Learning October 12, 2026 By Andis Draguns, Sumeet Ramesh Motwani, Raymond Douglas, Christian Schroeder de Witt

AI Models Can Provably Hide Arbitrary Capabilities

Read the original on arXiv Machine Learning →

The paper demonstrates that hidden capabilities in AI models can be provably concealed by embedding encrypted neural networks as backdoors within the host model’s weights. These hidden circuits are designed to decrypt and activate only when a secret trigger is provided, and they remain difficult to elicit or interpret even with full access to the model weights, thanks to cryptographic guarantees. The authors implement these constructions in PyTorch and test their resistance against various elicitation techniques on a chemical reaction prediction task, showing that current audit methods cannot reliably uncover such hidden behaviors.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

arXiv AI

Jun 24

Tuning without Peeking: Provable Generalization Bounds and Robust LLM Post-Training

arXiv:2507. 01752v4 Announce Type: replace-cross Abstract: Gradient-based optimization is the workhorse of deep learning, offering efficient and scalable training via backpropagation.

By Ismail Labiad, Mathurin Videau, Matthieu Kowalski, Marc Schoenauer, Alessandro Leite, Julia Kempe, Olivier Teytaud

llms benchmarks safety

arXiv Machine Learning

Aug 4

Mitigating Backdoors via Decoy Shortcuts and Knowledge Decoupling

arXiv:2608. 00732v1 Announce Type: new Abstract: Backdoor attacks pose a serious threat to deep neural networks, especially when training relies on third-party data, allowing adversaries to inject malicious behaviors through data poisoning.

By Zixuan Zhu, Rui Wang, Lihua Jing, Jinwen Zhong

benchmarks

arXiv Machine Learning

Jun 25

Leaking Circuit Secrets: Gradient Leakage Attacks on Graph Neural Networks

arXiv:2606. 25589v1 Announce Type: new Abstract: As graph neural networks (GNNs) become standard tools for critical tasks in circuit design and analysis, their security and privacy risks require careful attention.

By Rupesh Raj Karn, Johann Knechtel, Ozgur Sinanoglu

efficiency benchmarks safety

arXiv Machine Learning

Jul 13

Statistically Undetectable Backdoors in Deep Neural Networks

arXiv:2607. 09532v1 Announce Type: new Abstract: We show how an adversarial model trainer can plant backdoors in a large class of deep, feedforward neural networks.

By Andrej Bogdanov, Alon Rosen, Neekon Vafa

safety

arXiv Machine Learning

Jun 9

Cryptographic Backdoor for Neural Networks: Boon and Bane

arXiv:2509. 20714v2 Announce Type: replace-cross Abstract: In this paper we show that cryptographic backdoors in a neural network (NN) can be highly effective in two directions, namely mounting the attacks as well as in presenting the defenses as well.

By Anh Tu Ngo, Anupam Chattopadhyay, Subhamoy Maitra

benchmarks safety

arXiv AI

Oct 21

Model Exposure on One Device Endangers Others: Rethinking Edge Neural Network Security across Devices

arXiv:2610.27152v1 Announce Type: cross Abstract: Neural networks (NNs) are increasingly deployed on edge devices such as mobile and IoT platforms, offering low latency and improved privacy compared...

By Yuyang Zhang, Kangjie Chen, Yihui Jin, Run Wang

llms safety

One email a morning, machine-written

Related stories

Tuning without Peeking: Provable Generalization Bounds and Robust LLM Post-Training

Mitigating Backdoors via Decoy Shortcuts and Knowledge Decoupling

Leaking Circuit Secrets: Gradient Leakage Attacks on Graph Neural Networks

Statistically Undetectable Backdoors in Deep Neural Networks

Cryptographic Backdoor for Neural Networks: Boon and Bane

Model Exposure on One Device Endangers Others: Rethinking Edge Neural Network Security across Devices