AI 导读
X 用户 @kimmonismus 指出 OpenAI 在声明中用过去时描述暂停，其中一句写的是「this included a two-week pause」，据此认为为期两周的强化学习训练暂停已经结束，还引用「We wanted to take the time necessary」一句作为佐证。被引内容提到，OpenAI 暂停了最新部署模型的 RL 训练两周，规模最大的前沿 RL 训练仍处于搁置状态，期间只进行小规模训练和评估，以测试模型行为、安全防护和对齐证据。该决定源于初步发现其即将推出的 Astra 模型可能触及 OpenAI 的 Critical 网络安全阈值，以及与 Hugging Face 相关的事件。
正文
.@AndrewCurran_ He was paying very close attention again, and I didn't even notice. OpenAI writes in the past tense. So it seems the break is already over!
"(...) this included a two-week pause (...)"
"We wanted to take the time necessary (...)"
Not looking good for a soon GPT-Astra-release: OpenAI paused reinforcement learning on its latest deployment models for two weeks, and its largest planned frontier RL run remains on hold. The company is running smaller-scale training and evaluations while it tests model behavior, safeguards, and evidence of alignment. The decision follows preliminary findings that its upcoming Astra model may have reached OpenAI’s “Critical” cybersecurity threshold, alongside the OpenAI–Hugging Face incident. "While some Astra training and evaluations meet those requirements, a significant number of workloads remain paused until they are fully migrated and enhanced to meet the new security bar" Dont think Astra will be released any time soon. The question is: will china catch up in the meantime?在 X 查看被引用的帖子
来源：@kimmonismus · x.com