This is the preprint of the work has been accepted for publication at the IEEE Conference on Secure and Trustworthy Machine Learning (SaTML 2026). Cascading Robustness Verification: Toward Efficient Model‑Agnostic Certification

Mohammadreza Maleki Electrical, Computer, and Biomedical Engineering Toronto Metropolitan University Toronto, ON, Canada ORCID: 0000-0003-2871-8797 Rushendra Sidibomma Computer Science & Engineering University of Minnesota Twin-Cities Minneapolis, MN, USA ORCID: 0009-0009-5169-7569 Arman Adibi Compu · 2026

Read This is the preprint of the work has been accepted for publication at the IEEE Conference on Secure and Trustworthy Machine Learning (SaTML 2026). Cascading Robustness Verification: Toward Efficient Model‑Agnostic Certification by Mohammadreza Maleki Electrical, Computer, and Biomedical Engineering Toronto Metropolitan University Toronto, ON, Canada ORCID: 0000-0003-2871-8797 Rushendra Sidibomma Computer Science & Engineering University of Minnesota Twin-Cities Minneapolis, MN, USA ORCID: 0009-0009-5169-7569 Arman Adibi Compu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-04236.txt.

Description

One verifier's blind spot need not be everyone's. Mohammadreza Maleki and 3 co-authors propose Cascading Robustness Verification in a 2026 paper. Proving that a neural network resists attack examples is expensive, so users lean on fast but partial verifiers, each with its own shortcuts.

Their insight is simple: an input counts as certified if any method in the cascade certifies it. Chaining verifiers therefore raises measured robustness while keeping costs low. The method is model-blind, no matter how the network was trained. The authors also expose limits of the standard robustness metric itself, arguing the field has been missing certified safety.

Catalog record

CREATORMohammadreza Maleki Electrical, Computer, and Biomedical Engineering Toronto Metropolitan University Toronto, ON, Canada ORCID: 0000-0003-2871-8797 Rushendra Sidibomma Computer Science & Engineering University of Minnesota Twin-Cities Minneapolis, MN, USA ORCID: 0009-0009-5169-7569 Arman Adibi Compu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE69 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-04236

Browse the full library · About the Tower of Babel