And this way only LLMs can learn from these vulnerabilities, making it even more likely that the next one will be found by an LLM.
LLMs currently don’t learn, at least in the way most people mean, but I assume you mean the trainers will incorporate the knowledge in future releases.
LLMs currently don’t learn, at least in the way most people mean, but I assume you mean the trainers will incorporate the knowledge in future releases.