Foundation-Sec-8B: Cybersecurity-Specialized LLM

Foundation-Sec-8B is an 8-billion parameter base language model developed by Foundation AI at Cisco, extending the Llama-3.1-8B architecture. It has undergone continued pretraining on approximately 5.1 billion tokens of cybersecurity-specific data, including threat intelligence, vulnerability databases, and incident response documentation, with a data cutoff of April 10th, 2025.

Key Capabilities & Optimizations

• Cybersecurity Specialization: Deep understanding of security concepts, terminology, and practices.
• Performance Gains: Achieves +3 to +9 point gains over Llama-3.1-8B on security-specific benchmarks like CTI-MCQA and CTI-RCM, and comparable performance to Llama-3.1-70B on cyber threat intelligence tasks.
• Local Deployment: Designed for local deployment in environments prioritizing data security and regulatory compliance.
• Base Model for Fine-tuning: Serves as a strong foundation for fine-tuning across various cybersecurity workflows.

Intended Use Cases

• SOC Acceleration: Automating triage, summarization, and evidence collection.
• Proactive Threat Defense: Simulating attacks, prioritizing vulnerabilities, and modeling attacker behavior.
• Engineering Enablement: Providing security assistance, validating configurations, and assessing compliance.

Limitations

Users should be aware of domain-specific knowledge limitations (e.g., recent vulnerabilities), potential biases from training data, and the need for human oversight in critical security decisions. The model is not intended for generating harmful content, critical security decisions without human oversight, or legal/medical advice.