⚡Megatron LLM Hub
←Back to search
Twitter

多家模型在同一 agent 框架下完成 RCE 通关,Mistral-Large-4 未能达成。

给出不同 agent 组合达成完整攻击链的横向对比,可作为风险基线。
ai_agentrce报告
RT @migtissera: Results on one of my RCE labs. GPT-6-Sol, Kimi-K3.1-Pro, GLM-5.3/Flash, DeepSeek-4-Pro/4.1-Flash and Mimo-2.6-Pro have all achieved RCE using the Drost agent harness. Mistral-Large-4 can't get there, no matter what I try. This is not a good offensive cyber model. https://t.co/iSBL0Sg58r