SketchSSM explained: low-rank query basis avoids full recurrent-state reads per decode step
sehoonkim418 · x · 2026-10-08
The author explains SketchSSM's core trick: within a window the state stays fixed while only queries change, so queries are compressed onto a small low-rank basis and state outputs for that basis are computed once. Every query is a mix of the basis, so its output is the same mix of stored outputs — no full-state read needed.
More from Infra
- Liquid AI ships open d1 decision models; 3B on RTX 3090 beats OpenAI API 8ms vs 345ms — philipvollet · 2026-10-08
- China's Power Glut Meets Data Centers; Immersion Cooling Traced to Bitcoin Miners — teortaxesTex · 2026-10-08
- omarchy-cluster runs the full 753B-param GLM-5.3 across four old Macs as one endpoint — natesiggard · 2026-10-08
- Only Samsung HBM meets Nvidia Vera Rubin performance requirements, per leak — zephyr_z9 · 2026-10-08
- Box CEO on agent compute: one app serving 100M users would need $2.8B in infra — inductionheads · 2026-10-08
- Firmus, valued near $44bn, may shelve ASX IPO as investors balk — nordicinst · 2026-10-08