30 citations · 33 across the 10 of their papers we have counts for
11 papers
Does AI Help Cyber Attackers or Defenders? Evidence from Nonpublic Vulnerabilities and Subsequent Attacks
Tobias Heldt, Matt Turk, Christoph Landolt +1
The release decision for frontier AI systems increasingly relies on cyber capability benchmarks, yet public vulnerability benchmarks can expose agents to previously published advis…
The Oracle's Gambit: A Game-Theoretic Framework for Responsible AI Release
Christoph R. Landolt, Tobias Lorenz, Marta Kwiatkowska +1
Responsible vulnerability disclosure can secure the defender's head start by controlling when a vulnerability becomes public. However, this status quo is now challenged by increase…
Scalable Delphi: Large Language Models for Structured Risk Estimation
Tobias Lorenz, Mario Fritz
Quantitative risk assessment relies on structured expert elicitation to estimate unobservable properties. The Delphi method produces calibrated, auditable estimates but requires mo…
Certified Circuits: Stability Guarantees for Mechanistic Circuits
Alaa Anani, Tobias Lorenz, Bernt Schiele +2
Understanding how neural networks arrive at their predictions is essential for debugging, auditing, and deployment. Mechanistic interpretability pursues this goal by identifying ci…
Pixel-level Certified Explanations via Randomized Smoothing
Alaa Anani, Tobias Lorenz, Mario Fritz +1
Post-hoc attribution methods aim to explain deep learning predictions by highlighting influential input pixels. However, these explanations are highly non-robust: small, impercepti…
MIBP-Cert: Certified Training against Data Perturbations with Mixed-Integer Bilinear Programs
Tobias Lorenz, Marta Kwiatkowska, Mario Fritz
Data errors, corruptions, and poisoning attacks during training pose a major threat to the reliability of modern AI systems. While extensive effort has gone into empirical mitigati…
FullCert: Deterministic End-to-End Certification for Training and Inference of Neural Networks
Tobias Lorenz, Marta Kwiatkowska, Mario Fritz
Modern machine learning models are sensitive to the manipulation of both the training data (poisoning attacks) and inference data (adversarial examples). Recognizing this issue, th…
Adaptive Hierarchical Certification for Segmentation using Randomized Smoothing
Alaa Anani, Tobias Lorenz, Bernt Schiele +1
Certification for machine learning is proving that no adversarial sample can evade a model within a range under certain conditions, a necessity for safety-critical domains. Common…
Certified Robust Models with Slack Control and Large Lipschitz Constants
Max Losch, David Stutz, Bernt Schiele +1
Despite recent success, state-of-the-art learning-based models remain highly vulnerable to input changes such as adversarial examples. In order to obtain certifiable robustness aga…
Certifiers Make Neural Networks Vulnerable to Availability Attacks
Tobias Lorenz, Marta Kwiatkowska, Mario Fritz
To achieve reliable, robust, and safe AI systems, it is vital to implement fallback strategies when AI predictions cannot be trusted. Certifiers for neural networks are a reliable…
Speaking the Same Language: Matching Machine to Human Captions by Adversarial Training
Rakshith Shetty, Marcus Rohrbach, Lisa Anne Hendricks +2
While strong progress has been made in image captioning over the last years, machine and human captions are still quite distinct. A closer look reveals that this is due to the defi…