Building my own cyber model (defensive/offensive), testing a well known Chinese open source model so I don’t start from scratch and it seems is already good on offensive but not defensive, as showed in my testing Fascinating data. #cybersecurity #AI #LLM #SuperIntelligence