ClarusC64/liquidity-cascade-instability-v0.1

该数据集用于评估模型是否能够从简化的资产负债表和市场代理轨迹中检测金融流动性不稳定性。每个行代表一个简化的流动性压力场景，涵盖三个时间步。任务是判断系统是否保持稳定或趋向流动性级联。数据集的核心稳定性理念涉及现金储备轨迹、提款加速、信用利差轨迹、交易对手风险暴露、流动性设施响应、结算延迟和抵押品质量等多个因素的交互作用。预测目标是标签1表示流动性级联不稳定，标签0表示稳定或受控的流动性状态。数据集的结构包括多个变量，其中有些是干扰变量。该数据集是Clarus稳定性推理基准的一部分，支持金融系统中的稳定性推理研究。

This dataset evaluates whether models can detect financial liquidity instability from short balance-sheet and market-proxy trajectories. Each row represents a simplified liquidity-stress scenario across three time steps. The task is to determine whether the system remains stable or moves toward a liquidity cascade. The core stability idea involves the interaction reasoning across cash reserve trajectory, withdrawal acceleration, credit spread trajectory, counterparty exposure, liquidity facility response, settlement delay, and collateral quality. The prediction target is label = 1 for liquidity cascade instability and label = 0 for stable or controlled liquidity state. The dataset includes multiple variables, some of which are decoy variables. This dataset is part of the Clarus Stability Reasoning Benchmark and supports research into stability reasoning in financial systems.