Neural Memory Orchestration
Overseer is presented as an end-to-end learned page-prefetching system that combines an eBPF fault collector, a transformer predictor on network-attached inference hardware, and a kernel-resident enforcer. On two page-fault-heavy workloads, it is reported to cut runtime by roughly 35% for Hilbert-curve traversal and by about an order of magnitude for dense matrix multiplication, with fewer major faults.