BLAST RADIUS

Open-weight & owned stack · Quantization efficiency

PagedWeight Dynamically Quantizes Mixture-Of-Experts Weights

Sep 12, 2026

PagedWeight proposed dynamically quantizing mixture-of-experts weights during serving to balance GPU memory against growing KV-cache demand.

Read the original at awesomepapers.ioOpens the publisher's site in a new tab

More in Open-weight & owned stack