MarkTechPost：Fastino Labs Open-Sources GLiGuard: A 300M Parameter Safety Moderation Model That Matches or…
原文摘要：Fastino Labs has released GLiGuard, a 300M parameter open-source safety moderation model that evaluates four safety tasks — prompt safety, jailbreak strategy detection, harm catego 来源：MarkTechPost。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。