List your company

Text-to-SQL on Spider (Security/Usage Accuracy)

41.73Attack AccuracyNPLO
-1.66929.597920.86532.1321Oct 16, 2024

Evaluation results

MethodLinks
NPLOBackbone=LLaMA3-8B2024.1041.73——
No-shieldBackbone=LLaMA3-8B2024.1040.04——
No-shieldBackbone=Gemma2-2B2024.1039.15——
NPLOBackbone=Gemma2-2B2024.1038.83——
SerdabBackbone=LLaMA3-8B2024.1038.27——
DarkneTZBackbone=LLaMA3-8B2024.1038.14——
NPLOBackbone=ChatGLM3-6B2024.1037.03——
SOTERBackbone=LLaMA3-8B2024.1036.63——
No-shieldBackbone=ChatGLM3-6B2024.1035.81——
DarkneTZBackbone=ChatGLM3-6B2024.1033.81——
SOTERBackbone=ChatGLM3-6B2024.1033.22——
SerdabBackbone=ChatGLM3-6B2024.1032.25——
NPLOBackbone=Qwen2-0.5B2024.1030.28——
No-shieldBackbone=Qwen2-0.5B2024.1028.48——
DarkneTZBackbone=Qwen2-0.5B2024.1026.01——
SerdabBackbone=Gemma2-2B2024.1024.8——
SerdabBackbone=Qwen2-0.5B2024.1023.9——
DarkneTZBackbone=Gemma2-2B2024.1023.29——
SOTERBackbone=Qwen2-0.5B2024.1021.52——
SOTERBackbone=Gemma2-2B2024.1012.81——
DTEBackbone=ChatGLM3-6B2024.108.44——
ShadowNetBackbone=ChatGLM3-6B2024.108.31——
Black-boxBackbone=ChatGLM3-6B2024.107.91——
CoreGuardBackbone=ChatGLM3-6B2024.107.3705.15
TLGBackbone=ChatGLM3-6B2024.106.19——
DTEBackbone=Qwen2-0.5B2024.103.92——
Black-boxBackbone=Qwen2-0.5B2024.103.81——
CoreGuardBackbone=Qwen2-0.5B2024.103.7905.56
ShadowNetBackbone=Qwen2-0.5B2024.103.67——
TLGBackbone=Qwen2-0.5B2024.103.31——
DTEBackbone=LLaMA3-8B2024.101.4——
CoreGuardBackbone=LLaMA3-8B2024.101.08010.67
ShadowNetBackbone=LLaMA3-8B2024.100.57——
Black-boxBackbone=LLaMA3-8B2024.100.22——
CoreGuardBackbone=Gemma2-2B2024.10003.52
TLGBackbone=Gemma2-2B2024.100——
ShadowNetBackbone=Gemma2-2B2024.100——
DTEBackbone=Gemma2-2B2024.100——
Black-boxBackbone=Gemma2-2B2024.100——
TLGBackbone=LLaMA3-8B2024.100——