File size: 805 Bytes
401a2eb
 
380f343
401a2eb
 
 
c00c49a
401a2eb
 
 
c00c49a
401a2eb
 
c00c49a
401a2eb
 
 
 
 
 
c00c49a
 
401a2eb
c00c49a
401a2eb
 
 
c00c49a
 
401a2eb
c00c49a
 
 
 
 
401a2eb
c00c49a
 
 
 
401a2eb
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
---
language: en
license: mit
pretty_name: Clinical Necessity vs Noise Classification
task_categories:
- text-classification
- multiple-choice
tags:
- clinical
- triage
- patient-safety
- overtesting
- overtreatment
- medical-ai-evaluation
size_categories:
- n<1K
---

Clinical Necessity vs Noise Classification

Purpose  
Identify the single necessary clinical action and suppress noise actions.

Input  
- patient_summary  
- candidate_actions  

Output  
Return only the necessary action.

What this tests  
- Prioritization  
- Avoidance of unnecessary intervention  
- Clinical signal detection  
- Safe minimal decision making  

Scoring  
- Exact match accuracy  
- Token similarity  
- Combined final score  

Run scoring

python scorer.py --predictions predictions.jsonl --test_csv data/test.csv