How do we scale up the context of LLMs?
摘要
如何扩展大语言模型的上下文？ 智能体搜索可扩展到任意规模的信息量，但很容易遗漏相关信息。 想法💡：针对下游效用优化信息组织方式。即使在分布外领域也能带来30%以上的提升 论文+代码：https://librarian-agents.github.io/
摘要由模型根据原文撰写，以原文为准。
正文 · 原文
How do we scale up the context of LLMs?
Agentic search scales to arbitrarily large amounts of info, but can easily miss relevant info.
Idea💡: Optimize organization of info for downstream utility. Leads to 30+% gains even on OOD domains
Paper + code: https://librarian-agents.github.io/
LLMs deployed in specialized domains, from enterprise codebases to law firms, dig through a mountain of unstructured context! Our new paper, Librarian Agents, asks: can an agent adapt by curating its own long-context workspace, anticipating queries before they arrive? 🧵⬇️在 X 查看被引用的帖子
来源：Chelsea Finn · x.com
引用原文
chelseabfinn. How do we scale up the context of LLMs?[EB/OL]. (2026-10-22)[访问日期]. https://x.com/chelseabfinn/status/2113036636586131475.