COMPASS: Enhancing Agent Long-Horizon Reasoning with Evolving Context

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wan, Guangya, Ling, Mingyang, Ren, Xiaoqi, Han, Rujun, Li, Sheng, Zhang, Zizhao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909834830413824
author Wan, Guangya
Ling, Mingyang
Ren, Xiaoqi
Han, Rujun
Li, Sheng
Zhang, Zizhao
author_facet Wan, Guangya
Ling, Mingyang
Ren, Xiaoqi
Han, Rujun
Li, Sheng
Zhang, Zizhao
contents Long-horizon tasks that require sustained reasoning and multiple tool interactions remain challenging for LLM agents: small errors compound across steps, and even state-of-the-art models often hallucinate or lose coherence. We identify context management as the central bottleneck -- extended histories cause agents to overlook critical evidence or become distracted by irrelevant information, thus failing to replan or reflect from previous mistakes. To address this, we propose COMPASS (Context-Organized Multi-Agent Planning and Strategy System), a lightweight hierarchical framework that separates tactical execution, strategic oversight, and context organization into three specialized components: (1) a Main Agent that performs reasoning and tool use, (2) a Meta-Thinker that monitors progress and issues strategic interventions, and (3) a Context Manager that maintains concise, relevant progress briefs for different reasoning stages. Across three challenging benchmarks -- GAIA, BrowseComp, and Humanity's Last Exam -- COMPASS improves accuracy by up to 20% relative to both single- and multi-agent baselines. We further introduce a test-time scaling extension that elevates performance to match established DeepResearch agents, and a post-training pipeline that delegates context management to smaller models for enhanced efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2510_08790
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle COMPASS: Enhancing Agent Long-Horizon Reasoning with Evolving Context
Wan, Guangya
Ling, Mingyang
Ren, Xiaoqi
Han, Rujun
Li, Sheng
Zhang, Zizhao
Artificial Intelligence
Computation and Language
Long-horizon tasks that require sustained reasoning and multiple tool interactions remain challenging for LLM agents: small errors compound across steps, and even state-of-the-art models often hallucinate or lose coherence. We identify context management as the central bottleneck -- extended histories cause agents to overlook critical evidence or become distracted by irrelevant information, thus failing to replan or reflect from previous mistakes. To address this, we propose COMPASS (Context-Organized Multi-Agent Planning and Strategy System), a lightweight hierarchical framework that separates tactical execution, strategic oversight, and context organization into three specialized components: (1) a Main Agent that performs reasoning and tool use, (2) a Meta-Thinker that monitors progress and issues strategic interventions, and (3) a Context Manager that maintains concise, relevant progress briefs for different reasoning stages. Across three challenging benchmarks -- GAIA, BrowseComp, and Humanity's Last Exam -- COMPASS improves accuracy by up to 20% relative to both single- and multi-agent baselines. We further introduce a test-time scaling extension that elevates performance to match established DeepResearch agents, and a post-training pipeline that delegates context management to smaller models for enhanced efficiency.
title COMPASS: Enhancing Agent Long-Horizon Reasoning with Evolving Context
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2510.08790