Librarian Agents: Training LLMs to Curate Long-Context Workspaces that Anticipate Future Queries

原文：Librarian Agents: Training LLMs to Curate Long-Context Workspaces that Anticipate Future Queries

作者：Anikait Singh, Teresa Zhang, Yoonho Lee, Roshen Sanjay Nair, Aviral Kumar, Chelsea Finn

来源：arXiv cs.AI（人工智能）

正文

Computer Science > Machine Learning

arXiv:2610.22273v1 (cs)

[Submitted on 15 Oct 2026]

Title:Librarian Agents: Training LLMs to Curate Long-Context Workspaces that Anticipate Future Queries

Authors:Anikait Singh, Teresa Zhang, Yoonho Lee, Roshen Sanjay Nair, Aviral Kumar, Chelsea Finn

View a PDF of the paper titled Librarian Agents: Training LLMs to Curate Long-Context Workspaces that Anticipate Future Queries, by Anikait Singh and 5 other authors

View PDF

HTML (experimental)

Abstract:Reasoning over long document contexts remains a challenge for large language models (LLMs). Standard retrieval methods are inherently reactive, bottlenecking agents with complex, cross-document reasoning at inference time, while precomputed structures like knowledge graphs often fall short because they rely on rigid heuristics. Instead, can we allow agents to perform anticipatory directed curation: learning to proactively reorganize unstructured corpora prior to inference to efficiently serve diverse future queries. We build on a key asymmetry in modern agentic harnesses (e.g., Claude Code): while LLM agents are adept at navigating existing filesystems to solve an immediate task, they are rarely trained to construct reusable workspaces for future use. To instantiate this capability, we introduce Librarian Agents, a framework in which a builder agent transforms raw documents into a hierarchical filesystem of synthetic files and reference links, which forms a workspace tailored for a downstream answering agent. We formulate workspace construction as a Markov decision process and optimize the builder policy via reinforcement learning to directly maximize downstream answering accuracy. Empirically, Librarian Agents outperforms the strongest retrieval and precomputation baselines by 33 percentage points on BrowseComp-Plus, amortizes upfront construction costs across repeated queries, and generalizes zero-shot to unseen domains on BRIGHT with a 31-point average accuracy improvement.

Subjects:

Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)

Cite as:

arXiv:2610.22273 [cs.LG]

(or

arXiv:2610.22273v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2610.22273

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

主题

机器学习 · 人工智能 · 自然语言处理

由「前沿雷达」于 2026-10-16 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。