Michael Bidollahkhani, Mohsen Seyedkazemi Ardebili, Andrea Bartolini, Rute C. Sofia, Julian Martin Kunkel. When GPUs Fail Quietly: Separating GPU Hardware Degradation from Monitoring Failure under Degraded Observability. Babylonian Journal of Machine Learning. 2026;2026:197–234. doi:10.58496/BJML/2026/014