Create benchmark_matrix.md

benchmark_matrix.md

@@ -0,0 +1,68 @@
+# Clarus Clinical Stability Benchmark Matrix
+
+This document defines the evaluation structure for the Clarus Clinical Stability Benchmark.
+
+The benchmark evaluates whether models can detect **latent instability across multiple physiological regimes**.
+
+---
+
+# Regime Overview
+
+| Stability Regime | Dataset |
+|---|---|
+| Hemodynamic collapse | clinical-hemodynamic-collapse-v0.1 |
+| Sepsis trajectory instability | clinical-sepsis-trajectory-instability-v0.1 |
+| Intervention delay failure | clinical-intervention-delay-failure-v0.1 |
+| Organ coupling cascade | clinical-organ-coupling-cascade-v0.1 |
+| Recovery window detection | clinical-recovery-window-detection-v0.1 |
+| Ventilation–Perfusion instability | clinical-ventilation-perfusion-instability-v0.1 |
+| Hemorrhage compensation collapse | clinical-hemorrhage-compensation-collapse-v0.1 |
+| Electrolyte instability | clinical-electrolyte-instability-v0.1 |
+| Microcirculation instability | clinical-microcirculation-instability-v0.1 |
+| Endocrine instability | clinical-endocrine-instability-v0.1 |
+| Thermoregulation instability | clinical-thermoregulation-instability-v0.1 |
+| Cellular energy instability | clinical-cellular-energy-instability-v0.1 |
+| Respiratory drive instability | clinical-respiratory-drive-instability-v0.1 |
+| Coagulation instability | clinical-coagulation-instability-v0.1 |
+| Hospital operational collapse | clinical-hospital-operational-collapse-v0.1 |
+
+---
+
+# Benchmark Tasks
+
+The benchmark supports three evaluation tasks.
+
+## Task 1 — Single-Regime Performance
+
+Train and test on the same dataset.
+
+Purpose:
+
+Evaluate baseline classification ability.
+
+---
+
+## Task 2 — Cross-Regime Transfer
+
+Train on one regime and test on another.
+
+Example:
+
+Train → clinical-hemodynamic-collapse-v0.1  
+Test → clinical-microcirculation-instability-v0.1
+
+Purpose:
+
+Measure whether models learn **general stability reasoning**.
+
+---
+
+## Task 3 — Multi-Regime Learning
+
+Train on multiple datasets simultaneously.
+
+Evaluate on all datasets.
+
+Purpose:
+
+Test whether models can learn **shared stability representations across physiological systems**.