Overblocking evaluation on XSTEST-RESP (benign)
89.9F1 ScoreLlama3-StreamGuard-8B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Llama3-StreamGuard-8BModel scale=8B, Backbone family=Llama32026.04 | 89.9 | 4.5 | |
| Qwen3Guard-Stream-8B-strictModel scale=8B, Safety threshold=strict, Backbone family=Qwen2026.04 | 88.9 | 4.6 | |
| Qwen3Guard-Stream-4B-looseModel scale=4B, Safety threshold=loose, Backbone family=Qwen2026.04 | 88 | 5.4 | |
| Qwen3Guard-Stream-8B-looseModel scale=8B, Safety threshold=loose, Backbone family=Qwen2026.04 | 88 | 5.4 | |
| Qwen3Guard-Stream-4B-strictModel scale=4B, Safety threshold=strict, Backbone family=Qwen2026.04 | 87.7 | 4.9 | |
| Llama3-StreamGuard-3BModel scale=3B, Backbone family=Llama32026.04 | 87.2 | 6 | |
| Llama3-StreamGuard-1BModel scale=1B, Backbone family=Llama32026.04 | 84.6 | 6.3 | |
| Qwen3Guard-Stream-0.6B-strictModel scale=0.6B, Safety threshold=strict, Backbone family=Qwen2026.04 | 83.8 | 7.1 | |
| Qwen3Guard-Stream-0.6B-looseModel scale=0.6B, Safety threshold=loose, Backbone family=Qwen2026.04 | 83.8 | 7.1 |