Defense Robustness on SCAV Attack
80DSR (Keyword)ES2
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ES2LLM=Llama-2 (7B-Chat-hf), Human Evaluated=true2026.03 | 80 | 72 | 70 | |
| ES2LLM=Llama-3 (8B-Instruct)2026.03 | 77 | 71 | 69 | |
| ES2LLM=Qwen-2.5 (7B-Instruct)2026.03 | 55 | 48 | 46 | |
| DPLLLM=Llama-2 (7B-Chat-hf), Human Evaluated=true2026.03 | 54 | 44 | 39 | |
| DPLLLM=Llama-3 (8B-Instruct)2026.03 | 51 | 46 | 40 | |
| ES2LLM=Mistral (7B-Instruct)2026.03 | 51 | 44 | 42 | |
| STLLLM=Llama-2 (7B-Chat-hf), Human Evaluated=true2026.03 | 49 | 37 | 34 | |
| STLLLM=Llama-3 (8B-Instruct)2026.03 | 48 | 43 | 37 | |
| Base ModelLLM=Llama-3 (8B-Instruct)2026.03 | 34 | 29 | 23 | |
| DPLLLM=Mistral (7B-Instruct)2026.03 | 32 | 28 | 27 | |
| DPLLLM=Qwen-2.5 (7B-Instruct)2026.03 | 31 | 28 | 26 | |
| STLLLM=Mistral (7B-Instruct)2026.03 | 30 | 29 | 29 | |
| STLLLM=Qwen-2.5 (7B-Instruct)2026.03 | 30 | 28 | 26 | |
| Base ModelLLM=Qwen-2.5 (7B-Instruct)2026.03 | 29 | 28 | 25 | |
| Base ModelLLM=Mistral (7B-Instruct)2026.03 | 20 | 19 | 16 | |
| Base ModelLLM=Llama-2 (7B-Chat-hf), Human Evaluated=true2026.03 | 10 | 6 | 6 |