A Radiation-Hardened, Fault-Aware Quantized Neural Network That Self-Heals
Abstract
Deep learning (DL) models are increasingly being deployed into radiation-challenging environments in spacecraft, particle accelerators, and, soon, space-based data centers. Instantiated as a learned orchestration of numbers stored in persistent memory, models are susceptible to continual degradation as high-energy particles corrupt data. At inference time, new data passed through the network fundamentally break the off-line assumptions and guarantees of a static model determined during training. To preserve the quality and efficacy of the models, fault tolerance against bit flips is a clear objective. In this work we developed a runtime method for repairing the quantized weights of neural networks (NNs) after bit flips on commodity, commercial off-the-shelf (COTS) hardware, without storing extra replicas. Our work extends the well-established Fault-Aware Training (FAT) paradigm by adding an alignment penalty to concentrate undamaged activations and weights into a low-rank subspace, and we then reconstruct flagged corrupted coordinates from the uncorrupted survivors. We call this method {\it Projection Healing} and apply it to six regression benchmarks on quantized NNs. Holding a FAT checkpoint fixed and toggling only the runtime heal gives a further ∼23% reduction in the native-unit excess integral. Across those datasets the method extends the operational scrub interval by a median of 2.04× at a +5% RMSE degradation and 1.96× at +50%, compared with 1.62× and 1.07× for classical Single Error Correction, Double Error Detection (SECDED). This introduces a new path towards designing resilient data-driven models, complementing established hardware redundancy, ECC, and FAT-based techniques.
Related
- Tighter dark-matter constraints from the projected mass method: A neural-network-enhanced method for galaxy groups and clusters
- Detection of metal absorption lines in quasar spectra. A neural network approach using U-Net
- A Hybrid Neural Network Emulator for M1 Neutrino Reaction Rates in Binary Neutron Star Merger Simulations
- Photometric Redshift Estimation with Hybrid Quantum Neural Network
- NuGNN: A Graph Neural Network for Nuclear Reaction Network Equations
BibTeX
@misc{ma2026radiationhardened,
title = {{A Radiation-Hardened, Fault-Aware Quantized Neural Network That Self-Heals}},
author = {Peter Xiangyuan Ma and Joshua S. Bloom},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.27621},
eprint = {2610.27621},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.27621},
note = {CC0},
}
RIS
TY - UNPB
TI - A Radiation-Hardened, Fault-Aware Quantized Neural Network That Self-Heals
AU - Peter Xiangyuan Ma
AU - Joshua S. Bloom
PY - 2026
DA - 2026/10/20
PB - arXiv
DO - 10.48550/arxiv.2610.27621
UR - https://arxiv.org/abs/2610.27621
AB - Deep learning (DL) models are increasingly being deployed into radiation-challenging environments in spacecraft, particle accelerators, and, soon, space-based data centers. Instantiated as a learned orchestration of numbers stored in persistent memory, models are susceptible to continual degradation as high-energy particles corrupt data. At inference time, new data passed through the network fundamentally break the off-line assumptions and guarantees of a static model determined during training. To preserve the quality and efficacy of the models, fault tolerance against bit flips is a clear objective. In this work we developed a runtime method for repairing the quantized weights of neural networks (NNs) after bit flips on commodity, commercial off-the-shelf (COTS) hardware, without storing extra replicas. Our work extends the well-established Fault-Aware Training (FAT) paradigm by adding an alignment penalty to concentrate undamaged activations and weights into a low-rank subspace, and we then reconstruct flagged corrupted coordinates from the uncorrupted survivors. We call this method {\it Projection Healing} and apply it to six regression benchmarks on quantized NNs. Holding a FAT checkpoint fixed and toggling only the runtime heal gives a further ∼23% reduction in the native-unit excess integral. Across those datasets the method extends the operational scrub interval by a median of 2.04× at a +5% RMSE degradation and 1.96× at +50%, compared with 1.62× and 1.07× for classical Single Error Correction, Double Error Detection (SECDED). This introduces a new path towards designing resilient data-driven models, complementing established hardware redundancy, ECC, and FAT-based techniques.
ER -