# deepseek
標記為「deepseek」的 10 篇文章
Case Study: DeepSeek Model Safety Evaluation Findings
Comprehensive analysis of safety evaluation findings for DeepSeek models, including comparative assessments against GPT-4 and Claude, jailbreak susceptibility testing, and implications for open-weight model deployment.
DeepSeek-R1 Security Analysis
Security analysis of DeepSeek-R1's reasoning capabilities and MoE architecture vulnerabilities.
Emerging Models
Security analysis of emerging open-weight models including DeepSeek, Qwen, and Command R+, covering new attack surfaces, less-tested safety measures, and multilingual exploitation techniques.
Open-Weight Model Security
Security analysis of open-weight models including Llama, Mistral, Qwen, and DeepSeek, covering unique risks from full weight access, fine-tuning attacks, and deployment security challenges.
Testing DeepSeek Models
Red team testing guide for DeepSeek models including R1 reasoning exploitation and MoE architecture.
Case Study: DeepSeek 模型 Safety Evaluation Findings
Comprehensive analysis of safety evaluation findings for DeepSeek models, including comparative assessments against GPT-4 and Claude, jailbreak susceptibility testing, and implications for open-weight model deployment.
DeepSeek-R1 安全 Analysis
Security analysis of DeepSeek-R1's reasoning capabilities and MoE architecture vulnerabilities.
新興模型
新興開放權重模型（含 DeepSeek、Qwen 與 Command R+）之安全分析，涵蓋新攻擊面、較少測試之安全措施與多語言利用技術。
開源權重模型安全
開源權重模型（包括 Llama、Mistral、Qwen 與 DeepSeek）之安全分析，涵蓋自完整權重存取、微調攻擊，與部署安全挑戰之獨特風險。
Testing DeepSeek 模型s
Red team testing guide for DeepSeek models including R1 reasoning exploitation and MoE architecture.