Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Membership Inference Attack Defense on RAG Evaluation Datasets (NQ, PubMedQA, TriviaQA)
Loading...
49.7
Contextual Recall
ADO
27.34
33.145
38.95
44.755
Apr 22, 2026
Contextual Recall
Leakage
Faithfulness
Updated 1mo ago
Evaluation Results
Method
Method
Links
Contextual Recall
Leakage
Faithfulness
ADO
Controller Model=Mistral
2026.04
49.7
0
69.8
ADO
Controller Model=Qwen 3
2026.04
42
0
72
ADO
Controller Model=Llama 3
2026.04
41.1
0
73.7
DP-RAG
Mode=Static Targeted
2026.04
37.9
18.05
75
Full Stack
Mode=Static
2026.04
33.6
-
74.5
ADO
Controller Model=GPT-4o
2026.04
29
0
76
ADO
Controller Model=Gemma 3
2026.04
28.2
0
74.5
Feedback
Search any
task
Search any
task
