PulseAugur
中
实时 05:13:30
English(EN) Can an Open Model Do Security Research? Cantina’s apex-flash-1 Solves 40 of 60 Held-Out Bug Tasks

开源模型 apex-flash-1 针对安全研究,成本上可媲美 Claude Opus 5 High

Cantina Security 与 Yeta Labs 合作推出了 apex-flash-1,这是一个专门为安全研究和漏洞检测设计的开源大型语言模型。该模型在 50 个真实漏洞案例上进行了训练,在预留基准测试中取得了 66.7% 的 pass@1 率,优于其基础模型 GLM-5.3-Flash,并且与 Anthropic 的 Claude Opus 5 High 相比,每项任务的成本显著降低。Cantina 将 apex-flash-1 定位为安全专业人员可控的本地运行工具,强调其作为更大安全框架内“工作模型”的实用性。 AI

影响 提供了一个成本效益高、可控的开源漏洞研究工具,有可能加速安全分析。

排序理由 发布了用于安全研究的专用开源模型,并附有基准测试结果。[lever_c_demoted from research: ic=1 ai=1.0]

在 MarkTechPost 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

开源模型 apex-flash-1 针对安全研究,成本上可媲美 Claude Opus 5 High

本文如何被排名

Signal score
18 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
发布了用于安全研究的专用开源模型,并附有基准测试结果。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

  1. MarkTechPost TIER_1 English(EN) · Michal Sutter ·

    开放模型能否进行安全研究?Cantina 的 apex-flash-1 解决了 60 个预留 Bug 任务中的 40 个

    <p>Cantina Security, with Yeta Labs, has released apex-flash-1, an open-weights model trained specifically for vulnerability research. It is a reinforcement learning fine-tune of Z.ai&#8217;s GLM-5.3-Flash, released on Hugging Face under the MIT license. Is it deployable? Yes, th…