ClarusC64/clinical-nutritional-metabolic-instability-v0.1

该数据集用于评估模型是否能检测由代谢需求和营养供应不平衡引起的不稳定性。每个数据行代表一个简化的代谢调节场景，包含三个时间点的观察数据。任务目标是判断代谢供应是保持稳定还是趋向于代谢不稳定。代谢不稳定的核心概念是能量需求超过可用代谢缓冲时发生。数据集包含多种信号轨迹（如葡萄糖、酮、卡路里摄入等）和干扰变量。预测目标为标签1表示代谢营养不稳定，标签0表示稳定代谢调节。数据集是Clarus稳定性推理基准的一部分。

This dataset evaluates whether models can detect instability caused by imbalance between metabolic demand and nutritional supply. Each row represents a simplified metabolic regulation scenario observed across three time points. The task is to determine whether metabolic supply remains stable or is moving toward metabolic instability. Metabolic instability occurs when energy demand exceeds available metabolic buffering. The dataset includes various signal trajectories (e.g., glucose, ketone, caloric intake) and decoy variables. The prediction target is label 1 for metabolic nutritional instability and label 0 for stable metabolic regulation. This dataset is part of the Clarus Stability Reasoning Benchmark.