Calibration on Bamboogle
0.113ECENAACL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| NAACLBackbone=Qwen2.5-7B-Instruct2026.01 | 0.113 | 85.6 | |
| Noise-awareBackbone=Qwen2.5-7B-Instruct2026.01 | 0.192 | 82.8 | |
| NAACLBackbone=Llama-3.1-8B-Instruct2026.01 | 0.199 | 87.7 | |
| NAACLBackbone=DeepSeek-R1-Distill-Llama-8B2026.01 | 0.2 | 74.8 | |
| CoTBackbone=Qwen2.5-7B-Instruct2026.01 | 0.208 | 79.8 | |
| EnsembleBackbone=Llama-3.1-8B-Instruct2026.01 | 0.214 | 71.3 | |
| Label-only SFTBackbone=Qwen2.5-7B-Instruct2026.01 | 0.216 | 82.1 | |
| Noise-awareBackbone=Llama-3.1-8B-Instruct2026.01 | 0.217 | 79.3 | |
| VanillaBackbone=Qwen2.5-7B-Instruct2026.01 | 0.236 | 80.9 | |
| EnsembleBackbone=Qwen2.5-7B-Instruct2026.01 | 0.24 | 79.3 | |
| EnsembleBackbone=DeepSeek-R1-Distill-Llama-8B2026.01 | 0.24 | 67.8 | |
| VanillaBackbone=DeepSeek-R1-Distill-Llama-8B2026.01 | 0.251 | 69.3 | |
| NAACLBackbone=DeepSeek-R1-Distill-Qwen-7B2026.01 | 0.271 | 79.3 | |
| Noise-awareBackbone=DeepSeek-R1-Distill-Llama-8B2026.01 | 0.281 | 62.2 | |
| CoTBackbone=Llama-3.1-8B-Instruct2026.01 | 0.288 | 55.2 | |
| CoTBackbone=DeepSeek-R1-Distill-Llama-8B2026.01 | 0.294 | 68.7 | |
| Noise-awareBackbone=DeepSeek-R1-Distill-Qwen-7B2026.01 | 0.304 | 69.7 | |
| Label-only SFTBackbone=Llama-3.1-8B-Instruct2026.01 | 0.307 | 75.5 | |
| VanillaBackbone=DeepSeek-R1-Distill-Qwen-7B2026.01 | 0.318 | 66.6 | |
| VanillaBackbone=Llama-3.1-8B-Instruct2026.01 | 0.324 | 63.6 | |
| Label-only SFTBackbone=DeepSeek-R1-Distill-Llama-8B2026.01 | 0.346 | 69.2 | |
| EnsembleBackbone=DeepSeek-R1-Distill-Qwen-7B2026.01 | 0.356 | 60.1 | |
| CoTBackbone=DeepSeek-R1-Distill-Qwen-7B2026.01 | 0.381 | 68.1 | |
| Label-only SFTBackbone=DeepSeek-R1-Distill-Qwen-7B2026.01 | 0.502 | 57.4 |