[Ant Open-Source Large Model Security Framework SingProbe Now Compatible with 29 Open-Source Models] Golden Finance reports that on September 17, Jiemian News learned that Ant AI Security Lab has open-sourced the large model security framework SingProbe, along with its code, models, and evaluation benchmarks. Currently, this tool is compatible with 29 mainstream open-source models and can simultaneously identify unsafe content and potential risks of fabricated information during the generation phase.