Out-of-Distribution Detection on MNIST vs EMNIST (test)
96.1AUROCDIMS
Evaluation results
| Method | Links | |
|---|---|---|
| DIMSBackbone=LeNet, Hessian Approximation=low-rank, Parameter samples=100, friction coefficient=0.52026.05 | 96.1 | |
| RLABackbone=LeNet, Hessian Approximation=low-rank, Parameter samples=1002026.05 | 84.6 | |
| LINLABackbone=LeNet, Hessian Approximation=low-rank, Parameter samples=1002026.05 | 72.8 | |
| LABackbone=LeNet, Hessian Approximation=low-rank, Parameter samples=1002026.05 | 63.8 |