arXiv：cs.AI· Songtao Huang, Zhijie Zhong, Xiaohan He, Shuaiyu Zhang, Zhiyin Yu, Li Kang, Heng Zhou, Yutao Fan, Xiangchao Yan, Bo Zhang, Shiyang Feng·· 11 小时前
LongHarness：用状态化证据管理实现 100M-Token 语料上的智能体推理
LongHarness: Advancing Agentic Reasoning over 100M-Token Corpora via Stateful Evidence Management
AI 导读
模型无关的智能体框架 LongHarness 提出 Evidence-Conclusion Coevolution 机制，用未解决的结论引导探索、用新证据修正结论，并配合 Evidence Lifecycle Management 保持证据可追溯。
正文
Authors:Songtao Huang, Zhijie Zhong, Xiaohan He, Shuaiyu Zhang, Zhiyin Yu, Li Kang, Heng Zhou, Yutao Fan, Xiangchao Yan, Bo Zhang, Shiyang Feng
Abstract:Corpus-level long-context reasoning requires agents to acquire and integrate evidence across document collections that exceed their context windows. Existing agent harnesses enable incremental access to these corpora, but acquired evidence and evolving conclusions often remain entangled in a growing interaction history. This implicit state management can lead to two coupled failure modes: unbounded exploration, where retrieval expands the interaction history without resolving decision-relevant information gaps, and evidence-state drift, where previously acquired evidence or subsequent corrections fail to inform the current conclusion. We present LongHarness, a model-agnostic harness that makes the interaction between acquired evidence and evolving conclusions explicit. Its core mechanism, Evidence-Conclusion Coevolution, uses unresolved conclusions to guide exploration toward missing information, while newly acquired evidence strengthens or revises the conclusions that directed the search. LongHarness couples this process with Evidence Lifecycle Management, which keeps accumulated evidence current and traceable throughout the interaction. Together, these mechanisms focus retrieval on remaining information needs and keep conclusions aligned with current evidence. We evaluate LongHarness on OfficeQA Pro V1 and V2, AA-LCR, and LongBench v2 using GLM-5.3-Flash and Qwen3.8-Flash. Compared with direct reasoning using the same backbone, LongHarness improves GLM-5.3-Flash exact match on OfficeQA Pro from 15.8% to 60.2% on V1 and from 4.4% to 61.1% on V2. Against state-of-the-art agent harnesses, it surpasses Claude Code by 12.0 percentage points on V1 and Codex by 18.9 points on V2. With Qwen3.8-Flash, LongHarness reaches 61.7% on V1, outperforming Claude Code by 8.3 points and Codex by 12.0 points.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.13075 [cs.AI] |
| (or arXiv:2610.13075v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.13075 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Songtao Huang [view email]
[v1]
Fri, 9 Oct 2026 06:32:02 UTC (404 KB)
来源：arXiv：cs.AI · arxiv.org