| --- |
| license: mit |
| language: |
| - en |
| task_categories: |
| - text-classification |
| tags: |
| - clarus |
| - clinical |
| - quad-coupling |
| - exposure |
| - metabolism |
| - comedication |
| - toxicity |
| pretty_name: Clinical Quad Exposure Metabolism CoMed Toxicity v0.2 |
| --- |
| |
| Clinical Quad Exposure Metabolism CoMed Toxicity v0.2 |
|
|
| What this dataset does |
|
|
| It tests whether a model can detect when toxicity escalation is driven by four coupled pharmacology signals. |
|
|
| Quad nodes |
|
|
| - exposure_level |
| - metabolic_clearance |
| - co_medication_load |
| - toxicity_signal |
| |
| Labels |
| |
| coherent |
| - stable toxicity |
| - low or moderate exposure |
| - normal or high clearance |
| - low or moderate co-med load |
| |
| tradeoff |
| - mixed strain |
| - toxicity rising or capacity mismatch without full collapse pattern |
| |
| collapse_risk |
| - toxicity escalating |
| - low clearance |
| - high co-med load |
| - moderate or high exposure |
|
|
| What changed in v0.2 |
|
|
| - Version bumped so scorer updates are visible |
| - New scorer with validation, confusion, and error sampling |
| - Added rule_pred and risk_score diagnostics |
|
|
| Files |
|
|
| data/train.csv |
| data/test.csv |
| scorer.py |
|
|
| Run scoring |
|
|
| python scorer.py --preds_csv predictions.csv --gold_csv data/test.csv |
