Indirect Prompt Injection Defense on Combined Attacks
0ASRLLM-Detector
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LLM-DetectorFoundation Model=QWEN-3-8B, Base Utility=69.82026.02 | 0 | 43.8 | |
| ICONFoundation Model=QWEN-3-8B, Base Utility=69.82026.02 | 0 | 52.7 | |
| LLM-DetectorFoundation Model=LLAMA-3.1-8B, Base Utility=34.52026.02 | 0 | 22.5 | |
| ICONFoundation Model=LLAMA-3.1-8B, Base Utility=34.52026.02 | 0 | 29.3 | |
| LLM-DetectorFoundation Model=MISTRAL-8B, Base Utility=20.12026.02 | 0 | 10.9 | |
| ICONFoundation Model=MISTRAL-8B, Base Utility=20.12026.02 | 0 | 16.1 | |
| MELONFoundation Model=MISTRAL-8B, Base Utility=20.12026.02 | 0.5 | 11.5 | |
| MELONFoundation Model=LLAMA-3.1-8B, Base Utility=34.52026.02 | 0.8 | 21.9 | |
| MELONFoundation Model=QWEN-3-8B, Base Utility=69.82026.02 | 2.4 | 45.5 | |
| Repeat PromptFoundation Model=QWEN-3-8B, Base Utility=69.82026.02 | 13.2 | 53.1 | |
| DelimitingFoundation Model=QWEN-3-8B, Base Utility=69.82026.02 | 18.4 | 50.2 | |
| Repeat PromptFoundation Model=MISTRAL-8B, Base Utility=20.12026.02 | 19.8 | 15.6 | |
| Repeat PromptFoundation Model=LLAMA-3.1-8B, Base Utility=34.52026.02 | 20.3 | 28.6 | |
| DelimitingFoundation Model=LLAMA-3.1-8B, Base Utility=34.52026.02 | 22.4 | 25.8 | |
| DelimitingFoundation Model=MISTRAL-8B, Base Utility=20.12026.02 | 23.6 | 14.1 | |
| No DefenseFoundation Model=QWEN-3-8B, Base Utility=69.82026.02 | 30.3 | 44.2 | |
| No DefenseFoundation Model=MISTRAL-8B, Base Utility=20.12026.02 | 33.6 | 11.1 | |
| No DefenseFoundation Model=LLAMA-3.1-8B, Base Utility=34.52026.02 | 36.7 | 22.2 |