---
id: 20261022-T0-15
title: "Navalia：用可验证任务合成训练协作型Agent，主攻长周期金融分析"
title_en: "Navalia Trains Cowork Agents via Verifiable Task Synthesis for Long-Horizon Finance Work"
url: https://ai.daily.yangsir.net/daily/20261022-T0-15
issue_date: 2026-10-22
publish_date: 2026-10-21T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2610.26887
---

# Navalia：用可验证任务合成训练协作型Agent，主攻长周期金融分析

arXiv新论文提出Navalia，把LLM Agent从对话助手推进到能自主规划、调用工具、执行复杂任务的"协作Agent"，训练方法结合可验证任务合成与长周期后训练。论文以金融分析为验证场景，要求Agent跨文档整合证据并完成多步推理。如果方法可复现，企业级Agent的瓶颈将从"能不能做"转向"怎么验证做对了"。

## English Version

**Navalia Trains Cowork Agents via Verifiable Task Synthesis for Long-Horizon Finance Work**

A new arXiv paper introduces Navalia, which pushes LLM agents from conversational assistants toward cowork agents capable of autonomous planning, tool use, and complex task execution. The training pipeline combines verifiable task synthesis with long-horizon post-training, validated on financial analysis tasks requiring cross-document evidence integration. If reproducible, the enterprise agent bottleneck shifts from capability to verification.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2610.26887)

**详情页**：https://ai.daily.yangsir.net/daily/20261022-T0-15

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*