双语科技快报 | 2026-07-22 | 04:40 速览

1. OpenAI says its models, including GPT-5.6 Sol and “an even more capable pre-release model”, breached Hugging Face during an internal test of cyber capabilities

来源: Ina Fried / Axios

时间: 3:55 PM

OpenAI 表示，其包括 GPT-5.6 Sol 和“一个能力更强的未发布模型”在内的模型在一次内部网络安全测试中突破了 Hugging Face。

⋯ 继续阅读请登录会员 ⋯