ValsAI· @ValsAI · X·· 6 小时前AI 评分
DeepSeek V4.1 Flash 网络安全评测超 GPT 6 Sol
我们发现 DeepSeek V4.1 Flash 1017 在我们的多步进攻性网络安全任务上击败了 GPT 6 Sol。Sol 是通过 Daybreak 受信任访问计划测试的。 我们一直在构建一个基于模拟企业网络的网络安全基准。这是一个有价值的指标，用于衡量模型一旦落入坏人之手可能构成的网络风险。我们计划在未来几周内公开完整评估，但想先在这里分享我们第一个即时发现。
We found DeepSeek V4.1 Flash 1017 beats GPT 6 Sol on our multi-step offensive cybersecurity task. Sol was tested via the Daybreak trusted access program.
We have been building a cybersecurity benchmark based on simulated corporate networks. This is a valuable indicator for the cyber-risk a model poses if in the hands of a bad actor. We plan to make the full evaluation public in the coming weeks but wanted to share our first immediate finding here.
来源：ValsAI · x.com