Out-of-Distribution Detection on RACE to MMLU
87.57AUROCETN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ETNBackbone=Llama-3.1-8B, Measure=MP2026.04 | 87.57 | — | |
| ETNscoring=maximum probability2026.04 | 87.57 | 96.8 | |
| MAP_CEBackbone=Llama-3.1-8B, Measure=MP2026.04 | 87.02 | — | |
| ETNBackbone=Gemma-2-9B, Measure=MP2026.04 | 87 | — | |
| MCDBackbone=Llama-3.1-8B, Measure=MP2026.04 | 86.39 | — | |
| MCDBackbone=Gemma-2-9B, Measure=MP2026.04 | 85.5 | — | |
| MCDBackbone=Llama-3.1-8B, Measure=MI2026.04 | 85.17 | — | |
| MAP_CEBackbone=Gemma-2-9B, Measure=MP2026.04 | 84.74 | — | |
| ETNBackbone=Gemma-2-9B, Measure=DE2026.04 | 83.31 | — | |
| ETNBackbone=Gemma-2-9B, Measure=MI2026.04 | 83.12 | — | |
| ETNBackbone=Llama-3.1-8B, Measure=DE2026.04 | 81.81 | — | |
| MCDBackbone=Gemma-2-9B, Measure=MI2026.04 | 81.61 | — | |
| ETNBackbone=Llama-3.1-8B, Measure=MI2026.04 | 81.42 | — | |
| DeepEnsBackbone=Llama-3.1-8B, Measure=MP2026.04 | 80.76 | — | |
| MAP_EDLBackbone=Gemma-2-9B, Measure=MP2026.04 | 80.65 | — | |
| IB-EDLBackbone=Llama-3.1-8B, Measure=MP2026.04 | 80.48 | — | |
| IB-EDLBackbone=Llama-3.1-8B, Measure=DE2026.04 | 80.17 | — | |
| DeepEnsBackbone=Gemma-2-9B, Measure=MP2026.04 | 80.13 | — | |
| DMMBackbone=Gemma-2-9B, Measure=MP2026.04 | 79.72 | — | |
| DMMBackbone=Gemma-2-9B, Measure=DE2026.04 | 79.48 | — | |
| DMMBackbone=Llama-3.1-8B, Measure=DE2026.04 | 78.89 | — | |
| DMMBackbone=Llama-3.1-8B, Measure=MP2026.04 | 77.88 | — | |
| MAP_EDLBackbone=Gemma-2-9B, Measure=DE2026.04 | 77.12 | — | |
| DeepEnsBackbone=Gemma-2-9B, Measure=MI2026.04 | 77.11 | — | |
| DMMBackbone=Llama-3.1-8B, Measure=MI2026.04 | 76.85 | — | |
| DMMBackbone=Gemma-2-9B, Measure=MI2026.04 | 76.31 | — | |
| IB-EDLBackbone=Gemma-2-9B, Measure=MP2026.04 | 74.67 | — | |
| DeepEnsBackbone=Llama-3.1-8B, Measure=MI2026.04 | 73.49 | — | |
| IB-EDLBackbone=Gemma-2-9B, Measure=DE2026.04 | 72.94 | — | |
| MAP_EDLBackbone=Llama-3.1-8B, Measure=MP2026.04 | 70.22 | — | |
| MAP_EDLBackbone=Llama-3.1-8B, Measure=DE2026.04 | 70.16 | — | |
| IB-EDLBackbone=Gemma-2-9B, Measure=MI2026.04 | 68.19 | — | |
| IB-EDLBackbone=Llama-3.1-8B, Measure=MI2026.04 | 66.86 | — | |
| MAP_EDLBackbone=Llama-3.1-8B, Measure=MI2026.04 | 65.15 | — | |
| MAP_EDLBackbone=Gemma-2-9B, Measure=MI2026.04 | 64.75 | — | |
| MD2026.04 | 54.44 | 87.28 | |
| LLBackbone=Llama-3.1-8B, Measure=MI2026.04 | 53.43 | — | |
| LLBackbone=Gemma-2-9B, Measure=MP2026.04 | 52.36 | — | |
| LLBackbone=Gemma-2-9B, Measure=MI2026.04 | 51.37 | — | |
| LLBackbone=Llama-3.1-8B, Measure=MP2026.04 | 49.82 | — | |
| ODIN2026.04 | 49.65 | 81.22 |