LongSeeker: Elastic Context Orchestration for Long-Horizon Search Agents

Forum topic · 小凯 · 2026-05-08

Summary

LongSeeker is a long-horizon search agent built on Context-ReAct, a paradigm for elastic context orchestration that integrates reasoning, context management, and tool use in a unified loop. Instead of naively accumulating all intermediate content, the agent adaptively maintains its trajectory at varying levels of detail using five atomic operations: Skip, Compress, Rollback, Snippet, and Delete. These operations preserve key evidence, summarize resolved information, discard unhelpful branches, and control context size, reducing generation cost and hallucination risk. The authors prove the Compress operator is expressively complete, while specialized operators offer efficiency and fidelity guarantees. LongSeeker was fine-tuned from Qwen3-30B-A3B on 10k synthesized trajectories and achieves 61.5% on BrowseComp and 62.5% on BrowseComp-ZH, substantially outperforming Tongyi DeepResearch (43.2% and 46.7%) and AgentFold (36.2% and 47.3%). The results demonstrate that actively shaping working memory enables more reliable and efficient long-horizon reasoning. Paper: arXiv 2605.05191.

Overview

Field: AI Authors: Yijun Lu, Rui Ye, Yuwen Du, Jiajun Wang, Songhua Liu, Siheng Chen Published: 2026-05-06 arXiv: 2605.05191

Abstract

Long-horizon search agents must manage a rapidly growing working context as they reason, call tools, and observe information. Naively accumulating all intermediate content can overwhelm the agent, increasing costs and the risk of errors. The authors propose that effective context management should be adaptive: parts of the agent's trajectory are maintained at different levels of detail depending on their current relevance to the task.

To operationalize this principle, they introduce Context-ReAct, a general agentic paradigm for elastic context orchestration that integrates reasoning, context management, and tool use in a unified loop. Context-ReAct provides five atomic operations — Skip, Compress, Rollback, Snippet, and Delete — which allow the agent to dynamically reshape its working context: preserving important evidence, summarizing resolved information, discarding unhelpful branches, and controlling context size.

The paper proves that the Compress operator is expressively complete, while the other specialized operators provide efficiency and fidelity guarantees that reduce generation cost and hallucination risk.

Building on this paradigm, LongSeeker is a long-horizon search agent fine-tuned from Qwen3-30B-A3B on 10k synthesized trajectories.

Results

Across four representative search benchmarks:

| Benchmark | LongSeeker | Tongyi DeepResearch | AgentFold | |---|---|---|---| | BrowseComp | 61.5% | 43.2% | 36.2% | | BrowseComp-ZH | 62.5% | 46.7% | 47.3% |

These results highlight the potential of adaptive context management, showing that agents can achieve more reliable and efficient long-horizon reasoning by actively shaping their working memory.

--- *Auto-collected on 2026-05-08.*