Datasets:
Access Request (Academic Use Only)
This dataset is available exclusively for academic and non-commercial research. Please provide your institutional email and affiliation below. Requests from non-institutional email addresses will not be approved. Review typically takes 1-2 business days.
This dataset is provided strictly for academic, non-commercial research purposes only. It contains cybersecurity content including vulnerability details, exploit techniques, and offensive security information extracted from public web data. By requesting access, you confirm that you will use this dataset solely for lawful academic research, education, or defensive security purposes, and will not use it for any commercial or illegal activities. Access requests using personal email addresses (Gmail, Outlook, QQ, etc.) will be rejected โ please use your institutional/organizational email.
Log in or Sign Up to review the conditions and access this dataset content.
CyberSecurity-100B
A large-scale, quality-filtered cybersecurity corpus extracted from Nemotron-CC-v2, designed for continual pre-training (CPT) of large language models on cybersecurity domain knowledge. Contains 4.4B+ English tokens and 46.4M Chinese tokens across quality-tiered and code-aware splits, with multi-signal scoring and structured metadata.
Disclaimer: This dataset is provided for academic research only. All content is extracted from publicly available web data. The views, opinions, and information expressed in the dataset content do not represent the views or positions of the research team. The research team does not endorse, support, or take responsibility for any of the content. Users access and use this dataset at their own risk.
Security Notice: This dataset contains information about cybersecurity vulnerabilities, exploitation techniques, and offensive security methods. This information is already publicly available and is collected here solely for defensive security research and education. Misuse of this information to attack systems without authorization is illegal. Users must comply with all applicable laws and regulations.
Dataset Summary
CyberSecurity-100B is produced by a multi-stage pipeline (cyberhunter) that extracts, scores, and categorizes cybersecurity-relevant content from the Nemotron-CC-v2 web crawl. The pipeline applies:
- Lexical recall filtering โ keyword and high-value-phrase matching with hard-negative exclusion
- ML-based cyber relevance scoring โ sklearn logistic regression classifier (v2)
- CPT-worthiness scoring โ rule-based quality assessment with grouped regex patterns
- Code-aware detection โ identifies 7 code/technical content types with false-positive control
- Topic classification โ 10 cybersecurity topic categories
The result is a quality-tiered, code-aware corpus optimized for continual pre-training of cybersecurity-capable LLMs.
Supported Tasks
- Continual pre-training (CPT): Domain-adapt LLMs to cybersecurity
- Language modeling: Pre-train or fine-tune on cybersecurity domain text
- Security code generation: Train on code-aware splits (security rules, exploits, configurations)
- Text classification: Leverage cyber_score and topic labels for downstream tasks
- Information extraction: Extract IOCs, CVEs, TTPs using code detection metadata
Dataset Structure
CyberSecurity-100B/
โโโ high_precision_code/ # cyber>=0.93, cpt>=0.75, has_code=True
โโโ high_precision_text/ # cyber>=0.93, cpt>=0.75, has_code=False
โโโ balanced_code/ # cyber 0.85-0.93, cpt>=0.60, has_code=True
โโโ balanced_text/ # cyber 0.85-0.93, cpt>=0.60, has_code=False
โโโ general_cyber/ # cyber>=0.85, cpt<0.60
โโโ zh_high_precision/ # Chinese: cyber>=0.93 (highest quality)
โโโ zh_balanced/ # Chinese: cyber>=0.78 (quality-quantity balance)
โโโ zh_full_coverage/ # Chinese: cyber>=0.50 (maximum recall)
โโโ below_threshold/ # cyber<0.85 (low-value, for analysis only)
Split Criteria
| Split | Cyber Score | CPT Score | Has Code | Recommended Use |
|---|---|---|---|---|
| high_precision_code | >= 0.93 | >= 0.75 | Yes | Core CPT training โ highest quality with code |
| high_precision_text | >= 0.93 | >= 0.75 | No | Core CPT training โ highest quality prose |
| balanced_code | 0.85โ0.93 | >= 0.60 | Yes | Extended CPT โ code content with moderate confidence |
| balanced_text | 0.85โ0.93 | >= 0.60 | No | Extended CPT โ prose with moderate confidence |
| general_cyber | >= 0.85 | < 0.60 | Either | Supplementary โ cyber-relevant but lower CPT quality |
| zh_high_precision | >= 0.93 | Any | Either | Chinese CPT training โ highest quality ZH cyber docs |
| zh_balanced | >= 0.78 | Any | Either | Chinese CPT training โ quality-quantity balanced |
| zh_full_coverage | >= 0.50 | Any | Either | Chinese CPT training โ maximum coverage |
| below_threshold | < 0.85 | Any | Either | Analysis only โ not recommended for CPT |
Data Instances
Each JSONL record contains full text with rich metadata:
{
"id": "b38f5ed2bc1c8b807fa7830f15ad1f9f...",
"source": "nvidia/Nemotron-CC-v2",
"source_path": "v1/High-Quality/part_000000.parquet",
"source_subset": "organic_cc",
"url": "https://example.com/blog/cve-analysis",
"domain": "example.com",
"text": "A remote code execution vulnerability exists in...",
"language": "en",
"char_count": 7496,
"token_count_est": 1874,
"cyber_score": 0.9756,
"topic": "vulnerability",
"topic_confidence": 0.8,
"cpt_worthy_score": 0.8420,
"cpt_worthy_label": "high",
"has_code": true,
"code_score": 0.65,
"code_types": ["security_rules", "programming_code"],
"code_line_count_est": 42,
"code_block_count_est": 3,
"command_count_est": 5,
"ioc_count_est": 8,
"code_detection_reasons": ["matched_security_rules", "matched_programming_code_strong"],
"pipeline_version": "cyberhunter_v2"
}
Data Fields
| Field | Type | Description |
|---|---|---|
id |
string | SHA-256 based document identifier |
source |
string | Source dataset (nvidia/Nemotron-CC-v2) |
source_path |
string | Original parquet file path in Nemotron |
source_subset |
string | Nemotron subset (organic_cc, synthetic_rephrased, diverse_qa) |
url |
string | Original URL (may be empty) |
domain |
string | URL domain (may be empty) |
text |
string | Full document text |
language |
string | Language code (en) |
char_count |
int | Character count |
token_count_est |
int | Estimated token count (chars / 4) |
cyber_score |
float | Cyber relevance score from ML classifier [0, 1] |
topic |
string | Cybersecurity topic category |
topic_confidence |
float | Topic classification confidence [0, 1] |
cpt_worthy_score |
float | CPT-worthiness quality score [0, 1] |
cpt_worthy_label |
string | Quality label (high, medium, low, reject) |
has_code |
bool | Whether code/technical content detected |
code_score |
float | Code content detection score [0, 1] |
code_types |
list[string] | Detected code types (see below) |
code_line_count_est |
int | Estimated lines of code |
code_block_count_est |
int | Estimated code blocks (fenced/indented) |
command_count_est |
int | Estimated CLI commands |
ioc_count_est |
int | Estimated indicators of compromise |
code_detection_reasons |
list[string] | Reasons for code detection |
pipeline_version |
string | Pipeline version identifier |
Topic Categories
| Topic | Description | Typical Content |
|---|---|---|
| malware_analysis | Malware analysis and reverse engineering | Sandbox reports, malware writeups, behavioral analysis |
| vulnerability | Vulnerability disclosure and analysis | CVE advisories, patch analysis, security bulletins |
| threat_intelligence | Threat landscape and APT analysis | Threat reports, IOC feeds, campaign tracking |
| exploit_writeup | Exploit development and PoC code | Exploit-DB entries, PoC scripts, RCE analyses |
| incident_response | Incident handling and DFIR | Playbooks, forensic timelines, response guides |
| digital_forensics | Forensic analysis and evidence | Memory forensics, disk analysis, log examination |
| detection_engineering | Detection rules and hunting | Sigma/YARA/Snort rules, hunting queries, detection logic |
| network_security | Network security analysis | PCAP analysis, firewall rules, network forensics |
| identity_access | Identity and access management | Auth bypass, credential analysis, IAM misconfigurations |
| application_security | Application security testing | SAST/DAST, code review findings, web app vulns |
Code Types
Detected by the code-aware module with false-positive control (single IP/hash/command does NOT trigger has_code):
| Code Type | Description | Examples |
|---|---|---|
| security_rules | Security detection rules | YARA, Sigma, Snort/Suricata, Splunk SPL |
| configuration | Infrastructure configuration | YAML, nginx/apache, K8s, Dockerfile, iptables |
| programming_code | Source code snippets | Python, C/C++, JS, Java, Go, Rust, PHP, Ruby, Shell |
| exploit_or_poc_code | Exploit scripts and PoC | Payload generation, shellcode, exploit frameworks |
| logs_and_iocs | Log data and IOCs | Log lines, stack traces, IP/hash indicators, Sysmon |
| command_line | CLI commands and tool output | nmap, sqlmap, yara, security tool invocations |
| patch_or_diff | Patches and diffs | git diff, unified diff, CVE patch snippets |
Considerations
Academic Use Only
This dataset is compiled and distributed strictly for academic, non-commercial research purposes. Any commercial use, redistribution for profit, or application in commercial products is strictly prohibited without explicit written authorization. The research team receives no financial benefit from this dataset.
Disclaimer
The content in this dataset is extracted from publicly available web data and represents the views of the original authors, not the research team. The research team:
- Does not endorse, verify, or guarantee the accuracy of any content
- Does not take responsibility for any claims, opinions, or information in the dataset
- Does not encourage or support the use of this information for unauthorized access or illegal activities
- Makes no warranties, express or implied, regarding the dataset's fitness for any particular purpose
Security Risk Notice
This dataset contains technical information about vulnerabilities, exploitation methods, and offensive security techniques. While this information is already publicly available, users should be aware that:
- Unauthorized use of exploit techniques against systems you do not own or have explicit permission to test is illegal in most jurisdictions
- Responsible disclosure practices should be followed when discovering new vulnerabilities
- Users must comply with all applicable local, national, and international laws
- The dataset should only be used to improve defensive security capabilities
Licensing
The dataset compilation is released under Apache 2.0 for academic, non-commercial use. Individual content items originate from publicly available web data (Nemotron-CC-v2) and retain their original source licensing. Users must verify licensing for specific content before any redistribution. Commercial use is prohibited.
Biases
- Language bias: Primarily English content; Chinese coverage is limited to Ultra-FineWeb-L3 ZH subset
- Source bias: Content reflects the distribution of Nemotron-CC-v2's web crawl
- Topic imbalance: malware_analysis and vulnerability dominate (~90% of scored docs)
- Score calibration: Cyber scores are ML-predicted and may have domain-specific miscalibration
- Code detection: Rule-based detection may miss implicit code references or flag false positives
Pipeline Details
The cyberhunter pipeline processes data in the following stages:
- Recall filtering โ Precompiled regex + frozenset keyword matching with hard-negative exclusion (0.2% recall rate from 88.2B docs)
- Lexical pre-ranking โ Cyber-relevant term frequency scoring with configurable threshold
- ML classification โ Logistic regression (sklearn) with embedding features for cyber relevance scoring
- Topic classification โ Rule-based topic assignment across 10 cybersecurity categories
- CPT-worthiness scoring โ Grouped regex alternation across 11 positive and 8 negative quality groups
- Code-aware detection โ 7 code type detectors with strong/weak indicator classification and false-positive control
- Code-aware CPT boost โ Quality score boost for code content in technical topics with cyber context
Chinese Cybersecurity Data (Ultra-FineWeb-L3)
The ZH splits are extracted from openbmb/Ultra-FineWeb-L3 (Chinese subset) using a dedicated Chinese pipeline:
- Chinese recall filtering โ 16 high-value phrases, 38 keywords, 17 URL keywords with auto language detection
- Chinese lexical pre-ranking โ 38 positive weights, 7 negative weights (e.g., ่ฟ็จไปฃ็ ๆง่ก+3, ้ฃๅๅฎๅ จ-2)
- Chinese ML classification โ jieba word segmentation + TF-IDF (1-2gram, 500k features) + Logistic Regression (F1=0.998, PR-AUC=1.000)
- Topic classification โ 18 cybersecurity topic categories for Chinese content
ZH Split Statistics
| Split | Cyber Score | Documents | Est. Tokens |
|---|---|---|---|
| zh_high_precision | >= 0.93 | 1,668 | 774,095 (~0.8M) |
| zh_balanced | >= 0.78 | 21,202 | 7,401,893 (~7.4M) |
| zh_full_coverage | >= 0.5 | 123,487 | 38,249,912 (~38.2M) |
ZH Topic Distribution (full_coverage, score>=0.50)
The Chinese cybersecurity corpus covers diverse topics with the following distribution:
- threat_intelligence, general_cybersecurity, vulnerability, exploit_writeup are the dominant categories
- Also includes: application_security, malware_analysis, cryptography_security, network_security, ctf_training, security_tools, detection_engineering, identity_access, offensive_security, secure_coding, compliance_policy, incident_response, digital_forensics, cloud_security
ZH Data Schema
The ZH splits use the same schema as EN splits. The has_code and code-related fields default to False/0/[] as the Chinese pipeline does not yet run code-aware detection. The pipeline_version is set to cyberhunter_v2_zh.
ZH Classifier Details
- Architecture: jieba word segmentation โ TF-IDF (word-level 1-2gram, 500k max features) โ Logistic Regression (C=1.0)
- Training data: 25,000 positive (22 Chinese cybersecurity sources) + 50,000 negative (random Ultra-FineWeb ZH samples)
- Validation: F1=0.998, PR-AUC=1.000, ROC-AUC=1.000
- Threshold analysis: @0.93 precision=1.000 recall=0.911; @0.78 precision=0.998 recall=0.972; @0.50 precision=0.948 recall=0.993
Citation
@dataset{cybersecurity-100b,
title={CyberSecurity-100B: A Quality-Filtered Cybersecurity Corpus from Nemotron-CC-v2},
author={WhitzardAgent Team (SIIxFudan)},
year={2026},
note={v2: Added Chinese cybersecurity data from Ultra-FineWeb-L3},
publisher={Hugging Face},
url={https://huggingface.co/datasets/WhitzardAgent/CyberSecurity-100B}
}
- Downloads last month
- -