DualResearch：用于答案重构的熵门控双图检索
DualResearch: Entropy-Gated Dual-Graph Retrieval for Answer Reconstruction
AI总结：
针对深度研究框架的上下文污染与执行脆弱问题，提出DualResearch双图检索融合框架，通过熵门控融合语义与因果图证据，在HLE和GPQA上分别提升7.7%和6.06%的准确率。
摘要：
深度研究框架通过编排外部工具来执行复杂的、多步骤的科学推理，这超出了单个大型语言模型的固有极限。然而，它仍然存在上下文污染、证据支持薄弱和执行路径脆弱的问题。为了解决这些问题，我们提出了DualResearch，一个检索与融合框架，它通过联合建模两个互补的图来匹配工具密集型推理的认知结构：一个广度语义图，编码稳定的背景知识；以及一个深度因果图，捕获执行溯源。每个图都有一个层原生的相关性函数、用于广度的种子锚定语义扩散，以及用于深度的带有可靠性权重的因果语义路径匹配。为了调和它们的异质性和查询相关的不确定性，DualResearch将每层的路径证据转换为答案分布，并通过具有全局校准的熵门控规则在log空间中进行融合。该融合会提高更确定通道的权重，并放大一致性。作为深度研究系统的补充，DualResearch将冗长的多工具执行日志压缩为简洁的推理图，并且我们展示了它可以稳定且有效地重构答案。在科学推理基准HLE和GPQA上，DualResearch取得了有竞争力的性能。使用开源系统InternAgent的日志文件，其在HLE上的准确率提高了7.7%，在GPQA上提高了6.06%。
英文摘要：
The deep-research framework orchestrates external tools to perform complex, multi-step scientific reasoning that exceeds the native limits of a single large language model. However, it still suffers from context pollution, weak evidentiary support, and brittle execution paths. To address these issues, we propose DualResearch, a retrieval and fusion framework that matches the epistemic structure of tool-intensive reasoning by jointly modeling two complementary graphs: a breadth semantic graph that encodes stable background knowledge, and a depth causal graph that captures execution provenance. Each graph has a layer-native relevance function, seed-anchored semantic diffusion for breadth, and causal-semantic path matching with reliability weighting for depth. To reconcile their heterogeneity and query-dependent uncertainty, DualResearch converts per-layer path evidence into answer distributions and fuses them in log space via an entropy-gated rule with global calibration. The fusion up-weights the more certain channel and amplifies agreement. As a complement to deep-research systems, DualResearch compresses lengthy multi-tool execution logs into a concise reasoning graph, and we show that it can reconstruct answers stably and effectively. On the scientific reasoning benchmarks HLE and GPQA, DualResearch achieves competitive performance. Using log files from the open-source system InternAgent, its accuracy improves by 7.7% on HLE and 6.06% on GPQA.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
