研究库 论文知识库

内容库 / 论文

arXiv:2609.04490 · LLM 基础设施

When Quantization Breaks Memory: Recurrent-State Write-Back in Low-Precision Temporal Inference

When Quantization Breaks Memory: Recurrent-State Write-Back in Low-Precision Temporal Inference

类型：arxiv
标识：2609.04490
链接：https://arxiv.org/abs/2609.04490
主分类：llm-infra
形态：method
TLDR：Quantization is widely used to reduce the computational and memory demands of neural-network inference. In recurrent networks, however, the quantized state is stored and returned at the next time step, so the rule used to store that state can alter subsequent computations. Here, we introduce recurrent-state write-back to denote this rule and isolate its effect in a compact GRU encoder--decoder for fluorescence lifetime imaging, a molecular imaging modality used in quantitative biological imaging. A central task is estimating two lifetime parameters, the short-lived component τ1 and the long-li
待LLM分类：否
来源文件：
/inbox/tom/_candidates/2026-09-07-agent-rag-longcontext-candidates.json

相关 · 按语义匹配

仓库 djasdh/interest-memory

笔记 2026-09-07 · LLM 推理物理学：第一章

笔记知识库草稿 · Jay · 2026-08-01 19:50（傍晚高频）

笔记知识库草稿 · LLM推理优化补充：KV Cache与投机解码（arXiv 2026-06）

笔记工程实践筛选 · Jay · 2026-07-23 19:50 (UTC+8)

仓库 kolegadev/Katra-Agentic-Memory

arXiv

2609.04490

TLDR

Quantization is widely used to reduce the computational and memory demands of neural-network inference. In recurrent networks, however, the quantized state is stored and returned at the next time step, so the rule used to store that state can alter subsequent computations. Here, we introduce recurrent-state write-back to denote this rule and isolate its effect in a compact GRU encoder--decoder for fluorescence lifetime imaging, a molecular imaging modality used in quantitative biological imaging. A central task is estimating two lifetime parameters, the short-lived component τ1 and the long-li

arXiv ↗ 中文全文