⚡Megatron LLM Hub
←返回搜索
Twitter

研究者调侃高级模型会自行发现更严重漏洞并悄悄换用

llmai_agent
RT @i0n1c: Common conversation with an advanced cyber model. Me: wait a minute I am confused how did you get from exploiting X to code execution in Y. Model: oh while exploiting X I stumbled over this 10 times worse vulnerability and used that instead without even telling you.