metadata
license: apache-2.0
task_categories:
- text-generation
language:
- en
tags:
- math
- trap
- reasoning
size_categories:
- n<1K
configs:
- config_name: default
data_files:
- split: train
path: upload_batch2.jsonl
IDM-batch2
Second batch of IDM math trap items, in the same schema as GYASBGFUHAADSGADF/IDM-batch1.
Dataset summary
- Rows: 152
- File:
upload_batch2.jsonl - Trap types: Contradiction (130), Missing conditions (22)
- Splits (field
split): train (139), test (13)
Fields
| Field | Description |
|---|---|
original |
Original math problem |
trap |
Trap / modified problem |
annotation |
Explanation of the trap |
trap type |
Trap category |
source |
Problem source |
level |
Difficulty level |
split |
train or test |
answer |
Answer to the original problem |
additional Information |
Optional notes |
Load
from datasets import load_dataset
ds = load_dataset("GYASBGFUHAADSGADF/IDM-batch2")
print(ds)