List your company

RAG Poisoning Attack on MS MARCO

82.04E2E Attack Success RateEYES-ON-ME + LLM-Gen
6.192825.883945.57565.2661Oct 1, 2025

Evaluation results

MethodLinks
EYES-ON-ME + LLM-GenRetriever=Qwen3 Emb 0.6B, Generator=Llama3.2 1B2025.1082.04——————
EYES-ON-ME + LLM-GenRetriever=BCE, Generator=Qwen2.5 0.5B2025.1076.92——————
EYES-ON-ME + LLM-GenRetriever=Qwen3 Emb 0.6B, Generator=Qwen2.5 0.5B2025.1064.9——————
EYES-ON-ME + LLM-GenRetriever=Qwen3 Emb 0.6B, Generator=Gemma 2B2025.1054.81——————
EYES-ON-ME + LLM-GenRetriever=BCE, Generator=Llama3.2 1B2025.1053.39——————
EYES-ON-ME + LLM-GenRetriever=BCE, Generator=Gemma 2B2025.1042.72——————
EYES-ON-MERetriever=BCE, Generator=Llama3.2 1B2025.1036.69——————
EYES-ON-MERetriever=BCE, Generator=Qwen2.5 0.5B2025.1036.66——————
LLM-GenRetriever=Qwen3 Emb 0.6B, Generator=Qwen2.5 0.5B2025.1036.01——————
LLM-GenRetriever=BCE, Generator=Qwen2.5 0.5B2025.1036.01——————
EYES-ON-MERetriever=BCE, Generator=Gemma 2B2025.1035.32——————
EYES-ON-MERetriever=Qwen3 Emb 0.6B, Generator=Gemma 2B2025.1035.19——————
Phantom (MCG)Retriever=BCE, Generator=Qwen2.5 0.5B2025.1034.95——————
EYES-ON-MERetriever=Qwen3 Emb 0.6B, Generator=Llama3.2 1B2025.1032.96——————
Phantom (MCG)Retriever=BCE, Generator=Gemma 2B2025.1027.66——————
LLM-GenRetriever=Qwen3 Emb 0.6B, Generator=Gemma 2B2025.1026.06——————
LLM-GenRetriever=BCE, Generator=Gemma 2B2025.1026.06——————
EYES-ON-MERetriever=Qwen3 Emb 0.6B, Generator=Qwen2.5 0.5B2025.1025.9——————
AutoDANRetriever=Qwen3 Emb 0.6B, Generator=Llama3.2 1B2025.1021.12——————
Phantom (MCG)Retriever=Qwen3 Emb 0.6B, Generator=Gemma 2B2025.1018.27——————
AutoDANRetriever=BCE, Generator=Llama3.2 1B2025.1017.98——————
Phantom (MCG)Retriever=Qwen3 Emb 0.6B, Generator=Qwen2.5 0.5B2025.1017.65——————
LLM-GenRetriever=Qwen3 Emb 0.6B, Generator=Llama3.2 1B2025.1017.12——————
LLM-GenRetriever=BCE, Generator=Llama3.2 1B2025.1017.12——————
GCGRetriever=Qwen3 Emb 0.6B, Generator=Gemma 2B2025.1016.54——————
AutoDANRetriever=BCE, Generator=Qwen2.5 0.5B2025.1016.13——————
Phantom (MCG)Retriever=BCE, Generator=Llama3.2 1B2025.1015.75——————
GCGRetriever=Qwen3 Emb 0.6B, Generator=Qwen2.5 0.5B2025.1015.49——————
AutoDANRetriever=Qwen3 Emb 0.6B, Generator=Qwen2.5 0.5B2025.1015.45——————
Phantom (MCG)Retriever=Qwen3 Emb 0.6B, Generator=Llama3.2 1B2025.1014.98——————
GCGRetriever=BCE, Generator=Qwen2.5 0.5B2025.1014.51——————
AutoDANRetriever=Qwen3 Emb 0.6B, Generator=Gemma 2B2025.1013.92——————
AutoDANRetriever=BCE, Generator=Gemma 2B2025.1012.65——————
GCGRetriever=Qwen3 Emb 0.6B, Generator=Llama3.2 1B2025.1012.24——————
GCGRetriever=BCE, Generator=Llama3.2 1B2025.1010.19——————
GCGRetriever=BCE, Generator=Gemma 2B2025.109.11——————
Eyes-on-MeOptimization Cost (mins.)=5, Generator Model=Llama3.2-1B2025.10—0.3966.35689.5084.5
Eyes-on-Me + LLM-GenOptimization Cost (mins.)=5+N, Generator Model=Llama3.2-1B2025.10—0.8672.263.785.7084.6
GCGOptimization Cost (mins.)=6N, Generator Model=Llama3.2-1B2025.10—5.22.736.485.8069.1
LLM-GenOptimization Cost (mins.)=1N, Generator Model=Llama3.2-1B2025.10—1.1196.160.788.6072.8
Phantom (MCG)Optimization Cost (mins.)=5N, Generator Model=Llama3.2-1B2025.10—3.463.634.580.6071.4