Primus-FineWeb: Cybersecurity Text Corpus for LLM Training

A dataset constructed by filtering cybersecurity-related text from FineWeb, a refined version of Common Crawl. It was created by Trend Micro's AI Lab and last updated on August 9, 2025. The dataset uses a high-quality seed set of manually curated cybersecurity text to train a binary classifier for filtering.

Use Cases

• Training cybersecurity-specific language models based on the filtered text corpus.
• Fine-tuning LLMs for threat intelligence analysis based on cybersecurity-related text.
• Developing binary classifiers for cybersecurity content detection based on the described filtering methodology.
• Researching the distribution of cybersecurity topics within web-scale text collections.

Strengths

• Leverages a high-quality seed dataset (Primus-Seed) of manually curated cybersecurity text.
• Uses FineWeb, a refined and filtered version of the Common Crawl web corpus.
• Last updated on August 9, 2025, indicating recent maintenance.

Limitations

• Row count, file formats, and column-level documentation are unknown, which may limit suitability assessment.
• The dataset's size and specific temporal coverage are not provided.
• Data may reflect the biases inherent to the source web corpus (Common Crawl).

Provenance

Source

Trend Micro AI Lab (trendmicro-ailab)

Collection Method

Constructed by filtering cybersecurity text from FineWeb using a classifier trained on the Primus-Seed dataset.

Time Range

null

Freshness

Last updated 2025-08-09 10:17:43

Geography

null