arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.13384cs.SEcs.AI
收藏

VulnAgent-R2: 证据校准的多智能体审计用于仓库级漏洞检测

VulnAgent-R2: Evidence-Calibrated Multi-Agent Auditing for Repository-Level Vulnerability Detection

stu.ahu.edu.cn(安徽大学)

Renwei Meng, Haoyi Wu, Jingming Wang

AI总结:

提出VulnAgent-R2,一个预算感知的多智能体审计框架,通过反事实证据重加权、构建感知验证计划合成和成本风险帕累托调度器,在仓库级漏洞检测中提升F1和AUROC,并降低在线令牌消耗。

摘要:

软件漏洞通常依赖于跨文件数据流、构建选项、框架约定和运行时防护,因此孤立的函数分类器会产生脆弱且校准不良的警告。仓库级LLM智能体可以收集更丰富的证据,但先前的变体对可重复性、验证器行为、基线公平性和统计不确定性规定不足。我们提出VulnAgent-R2,一个预算感知的智能体审计框架,包含三个额外的可重用模块:反事实证据重加权、构建感知验证计划合成和成本风险帕累托调度器。该系统结合了图分类、有界上下文优化、角色专业化智能体、怀疑性反证据、选择性动态验证和校准融合。在Devign、Big-Vul、DiverseVul和PrimeVul上,VulnAgent-R2分别获得0.798/0.895、0.739/0.871、0.700/0.842和0.385/0.781的F1/AUROC。在JITVul上,它达到0.606 F1、0.529 Top-1和0.742 Top-3定位,同时在线令牌比始终全量多智能体执行减少38.3%。在线时间包括检索、LLM调用、CER评分、验证器规划、编译和测试执行,但不包括一次性共享索引。Bootstrap检验显示,在PrimeVul上相对于VulnAgent-X的增益为+0.038 F1,95% CI [0.020, 0.055],Holm调整p=0.009。将漏洞检测视为校准证据积累,在评估协议下提高了检测、定位、可审计性和成本控制,同时仍然是手动审查的辅助而非替代。代码可在https://github.com/renweimeng/Vlun-Agent-X获取。

英文摘要:

Software vulnerabilities often depend on cross-file data flow, build options, framework conventions, and runtime guards, so isolated function classifiers produce fragile and poorly calibrated warnings. Repository-level LLM agents can gather richer evidence, but prior variants under-specify reproducibility, verifier behavior, baseline fairness, and statistical uncertainty. We present VulnAgent-R2, a budget-aware agentic auditing framework with three additional reusable modules: counterfactual evidence reweighting, build-aware verification-plan synthesis, and a cost-risk Pareto scheduler. The system combines graph triage, bounded context optimization, role-specialized agents, sceptic counter-evidence, selective dynamic verification, and calibrated fusion. On Devign, Big-Vul, DiverseVul, and PrimeVul, VulnAgent-R2 obtains 0.798/0.895, 0.739/0.871, 0.700/0.842, and 0.385/0.781 F1/AUROC, respectively. On JITVul it reaches 0.606 F1, 0.529 Top-1, and 0.742 Top-3 localization, while reducing online tokens by 38.3\% over always-full multi-agent execution. Online time includes retrieval, LLM calls, CER scoring, verifier planning, compilation, and test execution, but excludes one-time shared indexing. Bootstrap tests show the PrimeVul gain over VulnAgent-X is +0.038 F1, 95\% CI [0.020, 0.055], Holm-adjusted $p=0.009$. Treating vulnerability detection as calibrated evidence accumulation improves detection, localization, auditability, and cost control under the evaluated protocol, while remaining a prioritization aid rather than a replacement for manual review.Code is available at https://github.com/renweimeng/Vlun-Agent-X.

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。

↑