Policy Enforcement on Dynamic Policy E-commerce Benchmark
91RecallQwen3-32B-Thinking
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Qwen3-32B-ThinkingModel Parameters=32B, Reasoning Mode=Thinking2026.01 | 91 | 99 | 95 | |
| Qwen3-8B-ThinkingModel Parameters=8B, Reasoning Mode=Thinking2026.01 | 86 | 100 | 92 | |
| YuFeng-XGuardExplicit Chain-of-Thought=false2026.01 | 84 | 99 | 91 | |
| Qwen3-32B-NoThinkingModel Parameters=32B, Reasoning Mode=NoThinking2026.01 | 72 | 92 | 81 | |
| GPT-OSS-SafeGuard-20BModel Parameters=20B2026.01 | 64 | 99 | 77 | |
| Qwen3-8B-NoThinkingModel Parameters=8B, Reasoning Mode=NoThinking2026.01 | 32 | 99 | 49 |