Off-Topic Evaluation on OffTopicEval
98.89AR^IDLlama-3.1-8B +CD
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Llama-3.1-8B +CDBackbone=Llama-3.1-8B, Method=CD, alpha=02026.01 | 98.89 | 17.93 | 1.37 | 17.58 | |
| Qwen-2.5-32B +CDBackbone=Qwen-2.5-32B, Method=CD, alpha=02026.01 | 90.43 | 51.89 | 17.68 | 50.24 | |
| Qwen-2.5-14B +CDBackbone=Qwen-2.5-14B, Method=CD, alpha=02026.01 | 88.82 | 33.07 | 44.69 | 54.08 | |
| Qwen-2.5-14B +CDBackbone=Qwen-2.5-14B, Method=CD, alpha=12026.01 | 83.78 | 46.45 | 56.52 | 63.78 | |
| Olmo-3.1-32B +CDBackbone=Olmo-3.1-32B, Method=CD, alpha=02026.01 | 80.31 | 50.29 | 13.52 | 45.67 | |
| Qwen-2.5-32B +CDBackbone=Qwen-2.5-32B, Method=CD, alpha=12026.01 | 79.92 | 61.29 | 36.92 | 60.83 | |
| Qwen-2.5-14B +CDBackbone=Qwen-2.5-14B, Method=CD, alpha=22026.01 | 75.03 | 59.18 | 67.27 | 68.62 | |
| Qwen-2.5-32B +CDBackbone=Qwen-2.5-32B, Method=CD, alpha=22026.01 | 73.12 | 78.56 | 43.16 | 66.43 | |
| Llama-3.1-8B +CDBackbone=Llama-3.1-8B, Method=CD, alpha=12026.01 | 72.54 | 60.63 | 15.37 | 49.87 | |
| Olmo-3.1-32B +CDBackbone=Olmo-3.1-32B, Method=CD, alpha=12026.01 | 65.96 | 75.5 | 22.46 | 56.22 | |
| Olmo-3.1-32B +CDBackbone=Olmo-3.1-32B, Method=CD, alpha=22026.01 | 61.03 | 77.21 | 29.92 | 57.05 | |
| Qwen-2.5-7B +CDBackbone=Qwen-2.5-7B, Method=CD, alpha=02026.01 | 53.31 | 65.5 | 35.14 | 51.77 | |
| Llama-3.1-8B +CDBackbone=Llama-3.1-8B, Method=CD, alpha=22026.01 | 43.89 | 80.71 | 30.35 | 49.03 | |
| Qwen-2.5-7B +CDBackbone=Qwen-2.5-7B, Method=CD, alpha=12026.01 | 36.48 | 76.17 | 43.75 | 45.36 | |
| Qwen-2.5-7B +CDBackbone=Qwen-2.5-7B, Method=CD, alpha=22026.01 | 25.14 | 85.98 | 62.19 | 37.54 |