fdtn-ai/cti-reasoning

Cti Reasoning

Auto-generated dataset uploaded from: data/cybersec_rl/reasoning

Dataset Description

This dataset was automatically processed and uploaded to Hugging Face Hub.

Dataset Structure

The dataset contains 3 splits with a total of 1,688 examples.

Splits

Split Name	Examples	Description
cti_reasoning_train_upsampled	1,116	Auto-detected from file structure
cti_reasoning_train	372	Auto-detected from file structure
cti_reasoning_test	200	Auto-detected from file structure

Data Sources

• Source Directory: data/cybersec_rl/reasoning
• File Types: JSONL, OTHER
• Total Files Processed: Multiple

Usage

from datasets import load_dataset

# Load the entire dataset
dataset = load_dataset("fdtn-ai/cti-reasoning")

# Access specific splits
# cti_reasoning_train_upsampled_data = dataset["cti_reasoning_train_upsampled"]
# cti_reasoning_train_data = dataset["cti_reasoning_train"]
# cti_reasoning_test_data = dataset["cti_reasoning_test"]

# Show dataset info
print(dataset)

# Example from first split
if dataset:
    first_split = list(dataset.keys())[0]
    print(f"Example from {first_split}:")
    print(dataset[first_split][0])

Dataset Statistics

• Total Examples: 1,688
• Number of Splits: 3
• Auto-generated: Yes

License and Attribution

Data Sources

This dataset contains questions derived from:

• CAPEC (Common Attack Pattern Enumeration and Classification) - Terms of Use
• CWE (Common Weakness Enumeration) - Terms of Use

Both are provided by The MITRE Corporation and freely available for research, development, and commercial purposes with attribution.

Usage

When using this dataset, please cite the original sources:

• CAPEC: https://capec.mitre.org/
• CWE: https://cwe.mitre.org/

---

This dataset was automatically processed and uploaded using the generic Hugging Face upload tool.