Title: Hound: Relation-First Knowledge Graphs for Complex-System Reasoning in Security Audits Show Chinese title

Title: 猎犬：用于安全审计中复杂系统推理的关系优先知识图谱

Abstract: Hound introduces a relation-first graph engine that improves system-level reasoning across interrelated components in complex codebases. The agent designs flexible, analyst-defined views with compact annotations (e.g., monetary/value flows, authentication/authorization roles, call graphs, protocol invariants) and uses them to anchor exact retrieval: for any question, it loads precisely the code that matters (often across components) so it can zoom out to system structure and zoom in to the decisive lines. A second contribution is a persistent belief system: long-lived vulnerability hypotheses whose confidence is updated as evidence accrues. The agent employs coverage-versus-intuition planning and a QA finalizer to confirm or reject hypotheses. On a five-project subset of ScaBench[1], Hound improves recall and F1 over a baseline LLM analyzer (micro recall 31.2% vs. 8.3%; F1 14.2% vs. 9.8%) with a modest precision trade-off. We attribute these gains to flexible, relation-first graphs that extend model understanding beyond call/dataflow to abstract aspects, plus the hypothesis-centric loop; code and artifacts are released to support reproduction. Show Chinese abstract

Abstract: Hound引入了一个以关系为中心的图引擎，该引擎在复杂代码库中相互关联的组件上提高了系统级推理能力。 该智能体设计了灵活的、分析师定义的视图，并使用紧凑的注释（例如，资金/价值流，认证/授权角色，调用图，协议不变式）来锚定精确检索：对于任何问题，它会加载恰好相关的代码（通常跨越多个组件），以便能够向外查看系统结构并深入到决定性的代码行。 第二个贡献是一个持久的信念系统：长期存在的漏洞假设，其置信度随着证据的积累而更新。 该智能体采用覆盖率与直觉相结合的规划和QA最终确定器来确认或拒绝假设。 在ScaBench[1]的五个项目子集上，Hound在基线LLM分析器上提高了召回率和F1值（微召回率31.2%对8.3%；F1 14.2%对9.8%），仅带来适度的精度损失。 我们认为这些提升来自于灵活的关系优先图，这些图将模型理解扩展到调用/数据流之外的抽象方面，加上以假设为中心的循环；代码和相关成果已发布以支持复现。