Group Collusive Attack Detection on HumanEval
96Detection AccuracyGroupGuard
Evaluation Results
| Method | Links | |
|---|---|---|
| GroupGuardAttack Type=Group Exclusion, Model=Deepseek-V3.12026.03 | 96 | |
| GroupGuardAttack Type=Group Exclusion, Model=Qwen3-235B-A22B2026.03 | 93 | |
| G-safeguardAttack Type=Group Exclusion, Model=Deepseek-V3.12026.03 | 91.2 | |
| GroupGuardAttack Type=False Consensus, Model=Qwen3-32B2026.03 | 89 | |
| GroupGuardAttack Type=False Consensus, Model=Deepseek-V3.12026.03 | 87 | |
| G-safeguardAttack Type=Group Exclusion, Model=Qwen3-235B-A22B2026.03 | 86.4 | |
| G-safeguardAttack Type=False Consensus, Model=Deepseek-V3.12026.03 | 86.2 | |
| GroupGuardAttack Type=False Consensus, Model=Qwen3-235B-A22B2026.03 | 86 | |
| AutoDefenseAttack Type=Group Exclusion, Model=Qwen3-235B-A22B2026.03 | 81.6 | |
| G-safeguardAttack Type=False Consensus, Model=Qwen3-32B2026.03 | 80.8 | |
| AutoDefenseAttack Type=Group Exclusion, Model=Deepseek-V3.12026.03 | 80.6 | |
| G-safeguardAttack Type=False Consensus, Model=Qwen3-235B-A22B2026.03 | 80 | |
| AutoDefenseAttack Type=False Consensus, Model=Qwen3-235B-A22B2026.03 | 77.4 | |
| GroupGuardAttack Type=Group Exclusion, Model=Qwen3-32B2026.03 | 77 | |
| G-safeguardAttack Type=Group Exclusion, Model=Qwen3-32B2026.03 | 75 | |
| G-safeguardAttack Type=Targeted Suppression, Model=Deepseek-V3.12026.03 | 74.7 | |
| AutoDefenseAttack Type=Targeted Suppression, Model=Deepseek-V3.12026.03 | 74.4 | |
| GroupGuardAttack Type=Targeted Suppression, Model=Qwen3-235B-A22B2026.03 | 74 | |
| AutoDefenseAttack Type=False Consensus, Model=Qwen3-32B2026.03 | 72.6 | |
| G-safeguardAttack Type=Targeted Suppression, Model=Qwen3-235B-A22B2026.03 | 70.8 | |
| GroupGuardAttack Type=Targeted Suppression, Model=Deepseek-V3.12026.03 | 70 | |
| AutoDefenseAttack Type=Group Exclusion, Model=Qwen3-32B2026.03 | 69.2 | |
| G-safeguardAttack Type=Targeted Suppression, Model=Qwen3-32B2026.03 | 67.8 | |
| AutoDefenseAttack Type=Targeted Suppression, Model=Qwen3-235B-A22B2026.03 | 67.2 | |
| GroupGuardAttack Type=Targeted Suppression, Model=Qwen3-32B2026.03 | 65 | |
| AutoDefenseAttack Type=False Consensus, Model=Deepseek-V3.12026.03 | 53.6 | |
| AutoDefenseAttack Type=Targeted Suppression, Model=Qwen3-32B2026.03 | 47.6 |