Ilia Shumailov

Architectural Backdoors in Neural Networks

2022 · arXiv (Cornell University)

Machine learning is vulnerable to adversarial manipulation. Previous literature has demonstrated that at the training stage attackers can manipulate data and data sampling procedures to control model behaviour. A common attack goal is to plant …

On the Limitations of Stochastic Pre-processing Defenses

2022 · arXiv (Cornell University)

Defending against adversarial examples remains an open problem. A common belief is that randomness at inference increases the cost of finding adversarial inputs. An example of such a defense is to apply a random transformation …

Wide Attention Is The Way Forward For Transformers?

2022 · arXiv (Cornell University)

The Transformer is an extremely powerful and prominent deep learning architecture. In this work, we challenge the commonly held belief in deep learning that going deeper is better, and show an alternative design approach that …

SEA: Shareable and Explainable Attribution for Query-based Black-box Attacks

2023 · arXiv (Cornell University)

Machine Learning (ML) systems are vulnerable to adversarial examples, particularly those from query-based black-box attacks. Despite various efforts to detect and prevent such attacks, ML systems are still at risk, demanding a more comprehensive approach …

The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections

2025 · arXiv (Cornell University)

How should we evaluate the robustness of language model defenses? Current defenses against jailbreaks and prompt injections (which aim to prevent an attacker from eliciting harmful knowledge or remotely triggering malicious actions, respectively) are typically …