AI & ML Breaks Assumption

Proves that logic and lookup-table (LUT) based neural networks are structurally more resilient to hardware bit-flips than standard architectures.

March 25, 2026

Original Paper

From Arithmetic to Logic: The Resilience of Logic and Lookup-Based Neural Networks Under Parameter Bit-Flips

Alan T. L. Bacellar, Sathvik Chemudupati, Shashank Nag, Allison Seigler, Priscila M. V. Lima, Felipe M. G. França, Lizy K. John

arXiv · 2603.22770

The Takeaway

It shifts the robustness paradigm from numerical precision to architectural primitives, showing that discrete Boolean lookups provide a superior accuracy-resilience trade-off. This has significant implications for deploying models in high-radiation (space) or safety-critical edge environments where hardware-level errors are frequent.

From the abstract

The deployment of deep neural networks (DNNs) in safety-critical edge environments necessitates robustness against hardware-induced bit-flip errors. While empirical studies indicate that reducing numerical precision can improve fault tolerance, the theoretical basis of this phenomenon remains underexplored. In this work, we study resilience as a structural property of neural architectures rather than solely as a property of a dataset-specific trained solution. By deriving the expected squared er