LongSeeker: Elastic Context Orchestration for Long-Horizon Search Agents
Paper: https://arxiv.org/abs/2605.05191 Authors: Yijun Lu, Rui Ye, Yuwen Du, Jiajun Wang, Songhua Liu, Siheng Chen Published: 2026-05-06 | Category: Agentic Search
Overview
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. This paper proposes 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.
Context-ReAct Paradigm
To operationalize this principle, the authors introduce Context-ReAct, a general agentic paradigm for elastic context orchestration that integrates reasoning, context management, and tool use in a unified loop. It provides five atomic operations:
- Skip – bypass unhelpful content
- Compress – summarize resolved information
- Rollback – discard unhelpful branches
- Snippet – preserve important evidence at reduced detail
- Delete – control context size
- Agentic Information Retrieval (arXiv 2410.09713)
- AceSearcher: Bootstrapping Reasoning and Search for LLMs via Reinforcement Learning (arXiv 2509.24193)
- Synergizing RAG and Reasoning: A Systematic Review (arXiv 2504.15909)
Theoretically, the Compress operator is proven to be *expressively complete*, while the specialized operators provide efficiency and fidelity guarantees that reduce generation cost and hallucination risk.
LongSeeker Agent
LongSeeker is a long-horizon search agent fine-tuned from Qwen3-30B-A3B on 10,000 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: agents can achieve more reliable and efficient long-horizon reasoning by actively shaping their working memory.
Key Takeaways
1. Architecture: The agentic paradigm treats retrieval frequency and strategy itself as a learnable object, beyond the classic cascade of retrieval, rerank, and generation. 2. Theory: Formal guarantees on context operators (expressive completeness of Compress) ground the design in more than heuristics. 3. Practical impact: Controlling working context reduces token cost and hallucination risk in deep-research style agents. 4. Open questions: Evaluation credibility, latency/cost constraints, safety on the open web, and cross-lingual/multimodal scaling remain open challenges.
> Original abstract: "Long-horizon search agents must manage a rapidly growing working context as they reason, call tools, and observe information... 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."