fdtn-ai/Foundation-Sec-8B-Instruct
Foundation AI at Cisco developed Foundation-Sec-8B-Instruct, an 8-billion parameter instruction-tuned language model built on the Meta Llama-3.1-8B architecture. This model is specialized for cybersecurity applications, leveraging prior training in security concepts and practices. It excels at instruction-following for tasks like SOC acceleration, proactive threat defense, and engineering enablement, designed for local deployment in security-sensitive environments. The model demonstrates significant gains over Llama-3.1-8B-Instruct on security-specific benchmarks and competitive performance against models like GPT-4o-mini.
No reviews yet. Be the first to review!