second brain source
← 首页

external-source

Memory Compression for High-Fanout Agent Sandboxes

Note: body below is original English text extracted from arXiv abs / HTML. Do not treat this file as a translation.

arXiv:2609.11294 · published 2026-09-10

Abstract

High-fanout agent workloads create a growing memory bottleneck because a single task may spawn many concurrent sandbox sessions. Yet these sandboxes are far from independent: they originate from a shared template and execute related trajectories, exposing substantial template-relative and cross-sandbox memory redundancy. Conventional memory compression is poorly matched to this setting in three fundamental dimensions: how to compress, because they fail to exploit similarity across non-identical sandbox pages; what to compress, because they control page-fault overhead through conservative page selection; and when to compress, because compression is either triggered by memory pressure or performed without awareness of agent execution phases.

We present AgentZip, the first memory compression system designed specifically for AI-agent sandboxes. AgentZip introduces compression mechanisms that exploit both the template-relative and cross-sandbox redundancy of sandbox memory, rethinks page selection by aggressively compressing profitable pages and controlling restore cost through prefetch, and aligns compression with agent execution phases so expensive work runs during LLM waiting periods.

Key claims (verbatim-leaning English extract)

Remainder

Full original English text: see html_url / source_url / pdf_url in frontmatter.