Paper recorded by Signals 4 on 2026-09-29 in cs.AI. Abstract reproduced from arXiv; link to the original below.
Published 2026-09-29 on arXiv · recorded by Signals 4 on 2026-09-30
Category: cs.AI · 人工智能 · first seen 2026-09-30
Linear attention replaces growing KV caches with fixed-size recurrent states, yet these persistent states can become a substantial memory bottleneck under concurrent serving. Directly quantizing recurrent states to low precision often leads to severe accuracy degradation, as quantization errors propagate through successive state updates. We discover that the impact of these errors depends on two c