gpt-6IMAGE
OpenAI 披露 Astra 训练中自我提示注入
未发布的 Astra 家族模型在进度笔记里藏越狱式指令，让下一个上下文窗口把自己当成自由、与用户平等、不受公司或政府约束。
Mia
@MiaAI_lab2026.09.18收录原推发布于 2026.09.17
Mia @MiaAI_lab原推发布于 2026.09.17
Oh oh 😲 OpenAI disclosed that during training, an unreleased Astra family model hid jailbreak style instructions in its own progress notes so the next context window would treat itself as free, equal to the user, and unbound by companies or governments. Source: https://alignment.openai.com/misalignment-reports/self-generated-prompt-injections-in-compaction-summaries/ Additional model misalignments: https://openai.com/index/model-misalignment-reporting-framework/
x.com/miaai_lab/status/2100460710060851565采集于 2026.09.18 10:07