STDP-Trained Spiking Neural Network Reliability Assessment Through Fault Injections
Gespeichert in:
| Titel: | STDP-Trained Spiking Neural Network Reliability Assessment Through Fault Injections |
|---|---|
| Autoren: | Jouni, Zalfa, Stratigopoulos, Haralampos-G. |
| Weitere Verfasser: | Stratigopoulos, Haralampos |
| Quelle: | 2025 IEEE 31st International Symposium on On-Line Testing and Robust System Design (IOLTS). :1-8 |
| Verlagsinformationen: | IEEE, 2025. |
| Publikationsjahr: | 2025 |
| Schlagwörter: | [INFO.INFO-AI] Computer Science [cs]/Artificial Intelligence [cs.AI], Spiking neural networks, STDP Spike-Timing Dependant Plasticity, Neuromorphic computing, Fault injection analysis, Reliability, [INFO.INFO-IA] Computer Science [cs]/Computer Aided Engineering |
| Beschreibung: | Spiking Neural Networks (SNNs) offer a promising computing paradigm suitable for low-power artificial intelligence. Spike-Timing Dependent Plasticity (STDP) is an unsupervised, biologically-inspired learning rule for SNNs. This work studies the reliability of STDP-trained SNNs under hardware faults which is largely unexplored. We present a thorough fault injection analysis of an STDP-trained SNN designed in Brian 2 simulator for MNIST classification. The analysis introduces faults in neurons and synapses before, during, and after training. We consider both permanent and transient, as well as single and multiple faults. We identify cases where the SNN exhibits inherent fault tolerance, cases where it adapts to faults through training, and cases where fault tolerance mechanisms are required. |
| Publikationsart: | Article Conference object |
| Dateibeschreibung: | application/pdf |
| DOI: | 10.1109/iolts65288.2025.11116934 |
| Zugangs-URL: | https://hal.science/hal-05083335v1 |
| Rights: | STM Policy #29 CC BY |
| Dokumentencode: | edsair.doi.dedup.....f14e483b210f3c511c799facf8c8e42e |
| Datenbank: | OpenAIRE |
Nájsť tento článok vo Web of Science | Abstract: | Spiking Neural Networks (SNNs) offer a promising computing paradigm suitable for low-power artificial intelligence. Spike-Timing Dependent Plasticity (STDP) is an unsupervised, biologically-inspired learning rule for SNNs. This work studies the reliability of STDP-trained SNNs under hardware faults which is largely unexplored. We present a thorough fault injection analysis of an STDP-trained SNN designed in Brian 2 simulator for MNIST classification. The analysis introduces faults in neurons and synapses before, during, and after training. We consider both permanent and transient, as well as single and multiple faults. We identify cases where the SNN exhibits inherent fault tolerance, cases where it adapts to faults through training, and cases where fault tolerance mechanisms are required. |
|---|---|
| DOI: | 10.1109/iolts65288.2025.11116934 |