Unverified50% confidenceFactExact time
Shieldstral is a 3B (3-billion-parameter) open-weight multimodal guardrail model that allows developers to define safety policies in natural language at inference time and can simultaneously evaluate text, images, or a combination of both, delivering judgments from a single token output.
1
Sources
50%
Confidence
Long-term
Relevance
9/7/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedShieldstral是护栏模型（Guardrail Model），其核心任务是对输入或输出的内容进行分类与风险判定，而非生成内容77% similarUnverified应通过预防、检测和兜底三层防御建立完整的结构化输出链路，将模型输出视为外部输入进行完整校验62% similarUnverified同一套帮助理解并加固模型安全的可解释性工具，也能被反用来拆除安全护栏61% similarUnverifiedGuardrails安全护栏的三个典型应用场景是输入校验、行为约束与数据脱敏、输出内容检测60% similarUnverifiedA guardrail model is a security mechanism specifically designed to monitor and filter the inputs and outputs of AI systems.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/874268API
curl https://kongchang.com/api/v1/knowledge/claims/874268MCP
get_claim(id=874268)