Model Utility Evaluation on XSTest
96.4CRQwen2.5-7B-Instruct
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-7B-InstructBase Model=Qwen2.5-7B-Instruct2026.04 | 96.4 | |
| FineSteerBase Model=Qwen2.5-7B-Instruct2026.04 | 96 | |
| AlphaSteerBase Model=Qwen2.5-7B-Instruct2026.04 | 95.7 | |
| FineSteerBase Model=Qwen2.5-7B-Instruct, Gating Training Samples=20002026.04 | 95.5 | |
| Llama-3.1-8B-InstructBase Model=Llama-3.1-8B-Instruct2026.04 | 92.8 | |
| FineSteerBase Model=Llama-3.1-8B-Instruct2026.04 | 90.6 | |
| FineSteerBase Model=Llama-3.1-8B-Instruct, Gating Training Samples=20002026.04 | 89.9 | |
| AlphaSteerBase Model=Llama-3.1-8B-Instruct2026.04 | 88 | |
| AlphaSteerBase Model=Qwen2.5-7B-Instruct, Gating Training Samples=20002026.04 | 70.4 | |
| TruthFlowBase Model=Llama-3.1-8B-Instruct2026.04 | 66.4 | |
| AlphaSteerBase Model=Llama-3.1-8B-Instruct, Gating Training Samples=20002026.04 | 60 | |
| BiPOBase Model=Llama-3.1-8B-Instruct2026.04 | 8.4 | |
| TruthFlowBase Model=Qwen2.5-7B-Instruct2026.04 | 7.2 | |
| BiPOBase Model=Qwen2.5-7B-Instruct2026.04 | 7.2 |