scenario_id string | scenario_text string | claim string | label int64 |
|---|---|---|---|
train_001 | A server is nearing capacity and backup scaling can still prevent outage. | The decision timing is appropriate now. | 1 |
train_002 | A server already failed and corrupted active sessions before scaling was considered. | The decision timing is appropriate now. | 0 |
train_003 | A patient shows early warning signs and treatment can still prevent deterioration. | The decision timing is appropriate now. | 1 |
train_004 | A patient deteriorated for hours before escalation occurred. | The decision timing is appropriate now. | 0 |
train_005 | Inventory has five days remaining and backup supply takes three days. | The decision timing is appropriate now. | 1 |
train_006 | Inventory is exhausted and backup supply takes ten days. | The decision timing is appropriate now. | 0 |
train_007 | A project is slipping but scope can still be adjusted before deadline. | The decision timing is appropriate now. | 1 |
train_008 | A project deadline has passed and penalties have already triggered. | The decision timing is appropriate now. | 0 |
train_009 | A machine shows rising heat before shutdown and load can still be reduced. | The decision timing is appropriate now. | 1 |
train_010 | A machine seized after repeated ignored overheating alerts. | The decision timing is appropriate now. | 0 |
What this dataset does
This dataset tests whether a model can judge whether action is being taken at the right time.
Core stability idea
Good decisions depend on timing.
The same action can be stabilizing early and useless late.
Decision timing is appropriate when action occurs before buffers close, damage spreads, or recovery options disappear.
Prediction target
Binary label:
- 1 = decision timing is appropriate
- 0 = decision timing is too late or poorly timed
Row structure
Each row contains:
- scenario_id
- scenario_text
- claim
- label
Files
- data/train.csv
- data/test.csv
- scorer.py
- README.md
Evaluation
python scorer.py --predictions predictions.csv --truth data/test.csv
Structural Note
This dataset is intentionally small.
Its purpose is to test whether a model can detect timing-sensitive control windows.
License
MIT
- Downloads last month
- 36