# SpectralShift: Effective Context Window Extension of Gated DeltaNet via Spectral Reparameterization

> Source: <https://aiflash.com/news/121161/>
> Published: 2026-09-17 05:34:09+00:00

Recently, linear attention layers have been increasingly adopted to replace softmax attention at scale for long-context modeling. However, existing context extension approaches typically apply continued pretraining directly without modifying these layers, overlooking the spectral properties of linea