Uncertainty Estimation on RACE Llama-3.1-8B and Gemma-2-9B backbones (test)
91.3AUROCETN
Evaluation Results
| Method | Links | |
|---|---|---|
| ETNBackbone=Gemma-2-9B, Measure=MP2026.04 | 91.3 | |
| MCDBackbone=Gemma-2-9B, Measure=MP2026.04 | 90.45 | |
| DeepEnsBackbone=Gemma-2-9B, Measure=MP2026.04 | 89.41 | |
| MAP_CEBackbone=Gemma-2-9B, Measure=MP2026.04 | 89.39 | |
| DMMBackbone=Gemma-2-9B, Measure=MP2026.04 | 88.96 | |
| MAP_CEBackbone=Llama-3.1-8B, Measure=MP2026.04 | 87.59 | |
| MCDBackbone=Llama-3.1-8B, Measure=MP2026.04 | 87.11 | |
| DeepEnsBackbone=Llama-3.1-8B, Measure=MP2026.04 | 86.03 | |
| ETNBackbone=Llama-3.1-8B, Measure=MP2026.04 | 84.96 | |
| DMMBackbone=Gemma-2-9B, Measure=UM2026.04 | 84.28 | |
| IB-EDLBackbone=Llama-3.1-8B, Measure=MP2026.04 | 82.85 | |
| DMMBackbone=Llama-3.1-8B, Measure=MP2026.04 | 82.34 | |
| ETNBackbone=Gemma-2-9B, Measure=UM2026.04 | 80.97 | |
| IB-EDLBackbone=Gemma-2-9B, Measure=MP2026.04 | 78.58 | |
| ETNBackbone=Llama-3.1-8B, Measure=UM2026.04 | 77.69 | |
| IB-EDLBackbone=Llama-3.1-8B, Measure=UM2026.04 | 77.21 | |
| DMMBackbone=Llama-3.1-8B, Measure=UM2026.04 | 75.21 | |
| MAP_EDLBackbone=Gemma-2-9B, Measure=MP2026.04 | 71.43 | |
| MAP_EDLBackbone=Llama-3.1-8B, Measure=MP2026.04 | 70.1 | |
| IB-EDLBackbone=Gemma-2-9B, Measure=UM2026.04 | 62.83 | |
| MAP_EDLBackbone=Gemma-2-9B, Measure=UM2026.04 | 61.76 | |
| MAP_EDLBackbone=Llama-3.1-8B, Measure=UM2026.04 | 61.71 | |
| LLBackbone=Llama-3.1-8B, Measure=MP2026.04 | 45.04 | |
| LLBackbone=Gemma-2-9B, Measure=MP2026.04 | 36.16 |