用结构化证据追踪去中心化金融风险，减少误报，提升监管可审计性。
DeXposure-Claw: An Agentic System for DeFi Risk Supervision
- 基于图时间序列模型预测未来风险网络，提供可解释的决策依据。
- 通过压力测试与确定性监控生成带归因的预警信号，降低误判率。
- 专为监管设计评估体系，支持可量化的干预效果验证。
去中心化金融（DeFi）使监管者面临快速演进、高度关联的信用风险。通用大模型代理在此场景中表现不佳：它们容易过度解读弱证据并建议高风险干预，而现有评估无法衡量由此产生的误报。我们提出 DeXposure-Claw，一种基于预测的智能监管系统，将大模型决策纳入结构化证据流：(1) DeXposure-FM 是一个图时间序列基础模型，用于预测未来的风险暴露网络；(2) 确定性监控器和压力情景将这些预测转化为类型化警报、归因信号与情景证据；(3) 数据健康与置信度门控机制在发出可审计的监管工单前控制升级。我们进一步构建了 DeXposure-Bench，一个六轴评估框架，其决策轴通过监管对齐的绝对损失真实值和明确的误干预率来评分。在五年每周真实数据上的实验充分验证了该系统有效性。代码已开源：https://github.com/EVIEHub/DeXposure-Claw。
原文摘要 · Abstract (English)
Decentralized finance exposes supervisors to fast-moving, networked credit risks. General-purpose LLM agents fit this setting poorly: they over-read weak evidence and recommend high-stakes interventions, while existing evaluations offer no regulator-aligned way to measure the resulting false alarms. We introduce DeXposure-Claw, a forecast-grounded agentic supervision system that routes LLM decisions through structured evidence: (1) DeXposure-FM, a graph time-series foundation model, forecasts future exposure networks; (2) deterministic monitors and stress scenarios then turn those forecasts into typed alerts, attribution signals, and scenario evidence; and (3) data-health and confidence gates constrain escalation before DeXposure-Claw emits auditable supervisory tickets with rationales. We further develop DeXposure-Bench, a six-axis evaluation harness, whose decision axis scores tickets against a regulator-aligned absolute-loss ground truth and an explicit false-intervention rate. Experiments on five years of weekly real data fully support our system. Code is at https://github.com/EVIEHub/DeXposure-Claw.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。