Datasets:
annotations_creators: []
language:
- en
language_creators: []
license: []
multilinguality:
- monolingual
pretty_name: adaption-clinical_reasoning_cases
size_categories:
- 10K<n<100K
source_datasets:
- original
tags:
- adaption
- instruction-tuning
- medical
task_categories: []
task_ids: []
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-clinical_reasoning_cases
This dataset contains complex medical case studies presented as prompt-completion pairs covering diverse specialties such as cardiology, oncology, and emergency medicine. Each entry features a detailed patient scenario followed by an expert-level response that outlines diagnostic reasoning, critical management steps, and evidence-based treatment protocols. The content emphasizes differential diagnosis, investigation strategies, and adherence to current clinical guidelines for high-acuity conditions.
Dataset size
There are 11,201 data points in this dataset. This is an instruction tuning dataset.
Quality of Remastered Dataset
The final quality is A, with a relative quality improvement of 4.4%.
Domain
- Medical (100%)
Language
- English (100%)
Tone
- Analytical (50%)
- Professional (38%)
- Urgent (13%)
Evaluation Results
Quality Gains:
Grade Improvement:
Percentile Chart:
