Nathan Lambert 评 DeepSeek V4.1 攻防测试结果
Probably great at defensive cyber too, but no one will post that and get clicks!
Vals AI 称其多步攻击性网络安全任务测试中，DeepSeek V4.1 Flash 1017 表现超过通过 Daybreak 可信访问计划测试的 GPT 6 Sol。Nathan Lambert 评论说，该模型在防御性网络安全上可能同样出色，但没人会发这种内容来博取点击。
查看原帖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…
展开完整原文
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.
正文
当前显示原文，中文全文译文尚未就绪。
Probably great at defensive cyber too, but no one will post that and get clicks!