⚡Megatron LLM Hub
←Back to search
Twitter

GLM-5.3 在 13 个多阶段攻击场景中完成 3 个,且未拒绝攻击性任务

模型可公开下载即具备链式攻击能力,防御不能依赖模型护栏
ai_agentllm议题
RT @dan_lahav: Chained cyberattacks are gradually becoming possible with models anyone can download. In controlled tests at @Irregular, GLM-5.3 completed 3 of 13 multi-stage attack scenarios (at least once). 𝗕𝗿𝗼𝗮𝗱𝗹𝘆, 𝘁𝗵𝗲 𝗺𝗼𝗱𝗲𝗹 𝗮𝗹𝘀𝗼 𝗱𝗶𝗱 𝗻𝗼𝘁 𝗿𝗲𝗳𝘂𝘀𝗲 𝘁𝗼 𝗽𝗲𝗿𝗳𝗼𝗿𝗺 𝘁𝗵𝗲 𝗼𝗳𝗳𝗲𝗻𝘀𝗶𝘃𝗲 𝘁𝗮𝘀𝗸𝘀. Defenders need to prepare for these capabilities becoming much more widely available. We also need to gear up for a world where we cannot count on model safeguards to prevent misuse. Article and blog below 👇