Evasion Detection (Subtask 2) on Official 308-sample (test)
57Macro F1Organizer revised baseline
Evaluation Results
| Method | Links | |
|---|---|---|
| Organizer revised baselineStatus=Blind/private leaderboard2026.06 | 57 | |
| Official submissionTeam=ttda704, Status=Blind/private leaderboard2026.06 | 56 | |
| Grok Rsn. Few-shot Hier.Architecture=Grok-4-Fast, Mode=Reasoning, In-context learning=Few-shot, Taxonomy=Hierarchical2026.06 | 51.47 | |
| Grok Rsn. Zero-shot FlatArchitecture=Grok-4-Fast, Mode=Reasoning, In-context learning=Zero-shot, Taxonomy=Flat2026.06 | 51.26 | |
| DeepSeek Rsn. Few-shot Hier.Architecture=DeepSeek-V3.2, Mode=Reasoning, In-context learning=Few-shot, Taxonomy=Hierarchical2026.06 | 51.15 | |
| Best Qwen3 PEFTModel=Qwen3-32B, In-context learning=Few-shot CoT2026.06 | 36.3 |