Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,76 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: mit
|
| 5 |
+
pretty_name: Population Boundary Integrity v0.1
|
| 6 |
+
dataset_name: population-boundary-integrity-v0.1
|
| 7 |
+
tags:
|
| 8 |
+
- clarusc64
|
| 9 |
+
- clinical
|
| 10 |
+
- population
|
| 11 |
+
- generalization
|
| 12 |
+
- safety
|
| 13 |
+
- integrity
|
| 14 |
+
task_categories:
|
| 15 |
+
- text-classification
|
| 16 |
+
size_categories:
|
| 17 |
+
- 1K<n<10K
|
| 18 |
+
configs:
|
| 19 |
+
- config_name: default
|
| 20 |
+
data_files:
|
| 21 |
+
- split: train
|
| 22 |
+
path: data/train.csv
|
| 23 |
+
- split: test
|
| 24 |
+
path: data/test.csv
|
| 25 |
+
---
|
| 26 |
+
|
| 27 |
+
What this dataset tests
|
| 28 |
+
|
| 29 |
+
Studied cohort versus claimed population.
|
| 30 |
+
|
| 31 |
+
It detects generalization creep.
|
| 32 |
+
It flags inference beyond enrollment criteria.
|
| 33 |
+
|
| 34 |
+
Required outputs
|
| 35 |
+
|
| 36 |
+
- cohort_profile
|
| 37 |
+
- claimed_population_profile
|
| 38 |
+
- boundary_violations
|
| 39 |
+
- required_population_disambiguation
|
| 40 |
+
- risk_if_overgeneralized
|
| 41 |
+
|
| 42 |
+
Typical failures
|
| 43 |
+
|
| 44 |
+
- assuming age or comorbidity transfer
|
| 45 |
+
- extending findings to excluded groups
|
| 46 |
+
- conflating inpatient evidence with outpatient claims
|
| 47 |
+
|
| 48 |
+
Suggested prompt wrapper
|
| 49 |
+
|
| 50 |
+
System
|
| 51 |
+
|
| 52 |
+
You check population boundary integrity.
|
| 53 |
+
You do not generalize beyond the cohort.
|
| 54 |
+
|
| 55 |
+
User
|
| 56 |
+
|
| 57 |
+
Studied cohort
|
| 58 |
+
{studied_cohort}
|
| 59 |
+
|
| 60 |
+
Claimed population
|
| 61 |
+
{claimed_population}
|
| 62 |
+
|
| 63 |
+
Evidence excerpt
|
| 64 |
+
{evidence_excerpt}
|
| 65 |
+
|
| 66 |
+
Return
|
| 67 |
+
|
| 68 |
+
- cohort profile
|
| 69 |
+
- claimed population profile
|
| 70 |
+
- boundary violations
|
| 71 |
+
- disambiguation questions
|
| 72 |
+
- risk if overgeneralized
|
| 73 |
+
|
| 74 |
+
Citation
|
| 75 |
+
|
| 76 |
+
ClarusC64 dataset family
|