Sakana AI 发布 Fugu-Cyber：CyberGym 86.9%，CTI-REALM 72.1%

精选理由

Sakana AI 新出的网络安全模型 Fugu-Cyber 在 CyberGym 上拿了 86.9%，比 GPT-5.5-Cyber 还高，做安全测试的可以关注。

AI 摘要

Sakana AI 发布了基于 Fugu 编排模型的安全调优端点 Fugu-Cyber。该模型在 CyberGym 基准上取得 86.9% 准确率，在 CTI-REALM 上达到 72.1%。其性能超过了 GPT-5.5-Cyber 和 Claude Mythos Preview。访问需要手动批准并遵循防御性使用政策及 Token 计划。

Sakana AI 发布 Fugu-Cyber：CyberGym 86.9%，CTI-REALM 72.1% — 图片来源 · marktechpost

原文 · marktechpost

Sakana AI Releases Fugu-Cyber: An Orchestration Model Reporting 86.9% on CyberGym and 72.1% on CTI-REALM

Sakana AI has released Fugu-Cyber, a security-tuned endpoint on its Fugu orchestration model. It reports 86.9% on CyberGym and 72.1% on CTI-REALM, edging past GPT-5.5-Cyber and Claude Mythos Preview. Access is gated behind manual approval, a defensive-use policy, and the Token Plan. Here is what the numbers actually mean. The post Sakana AI Releases Fugu-Cyber: An Orchestration Model Reporting 86.9% on CyberGym and 72.1% on CTI-REALM appeared first on MarkTechPost .

阅读原文