# Image Classification on CIFAR-10 (test) (L1 ECEKDE Metric) Source: [Image Classification on CIFAR-10 (test) (L1 ECEKDE Metric)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:47:57.082Z Evaluation of Image Classification model performance on the CIFAR-10 test set using the L1 Expected Calibration Error with Kernel Density Estimation (L1 ECEKDE) metric. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Image Classification | | Dataset (including any stated split) | CIFAR-10 (test) | | Metric | L1 ECEKDE | | Unit | SCALAR | | Direction | Lower is better | | Results loaded | 32 of 32 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [L1 ECEKDE](): Lower is better; unit: SCALAR. ## Recorded results | Method | L1 ECEKDE | Result date | Report paper | | --- | --- | --- | --- | | L1 KDE-XE | 0.077 | Not recorded | 68fa6daa1e209131cea3b507 | | FL-53 | 0.081 | Not recorded | 68fa6daa1e209131cea3b507 | | FL-53 | 0.081 | Not recorded | 68fa6daa1e209131cea3b507 | | XE | 0.083 | Not recorded | 68fa6daa1e209131cea3b507 | | L1 KDE-XE | 0.09 | Not recorded | 68fa6daa1e209131cea3b507 | | L1 KDE-XE | 0.091 | Not recorded | 68fa6daa1e209131cea3b507 | | XE | 0.098 | Not recorded | 68fa6daa1e209131cea3b507 | | XE | 0.104 | Not recorded | 68fa6daa1e209131cea3b507 | | L1 KDE-XE | 0.124 | Not recorded | 68fa6daa1e209131cea3b507 | | L1 KDE-XE | 0.126 | Not recorded | 68fa6daa1e209131cea3b507 | | XE | 0.131 | Not recorded | 68fa6daa1e209131cea3b507 | | XE | 0.132 | Not recorded | 68fa6daa1e209131cea3b507 | | XE | 0.133 | Not recorded | 68fa6daa1e209131cea3b507 | | MMCE | 0.133 | Not recorded | 68fa6daa1e209131cea3b507 | | FL-53 | 0.138 | Not recorded | 68fa6daa1e209131cea3b507 | | MMCE | 0.143 | Not recorded | 68fa6daa1e209131cea3b507 | | FL-53 | 0.147 | Not recorded | 68fa6daa1e209131cea3b507 | | MMCE | 0.155 | Not recorded | 68fa6daa1e209131cea3b507 | | MMCE | 0.155 | Not recorded | 68fa6daa1e209131cea3b507 | | FL-53 | 0.156 | Not recorded | 68fa6daa1e209131cea3b507 | | L1 KDE-XE | 0.163 | Not recorded | 68fa6daa1e209131cea3b507 | | MMCE | 0.164 | Not recorded | 68fa6daa1e209131cea3b507 | | XE | 0.17 | Not recorded | 68fa6daa1e209131cea3b507 | | MMCE | 0.171 | Not recorded | 68fa6daa1e209131cea3b507 | | FL-53 | 0.171 | Not recorded | 68fa6daa1e209131cea3b507 | | XE | 0.172 | Not recorded | 68fa6daa1e209131cea3b507 | | FL-53 | 0.177 | Not recorded | 68fa6daa1e209131cea3b507 | | L1 KDE-XE | 0.178 | Not recorded | 68fa6daa1e209131cea3b507 | | MMCE | 0.196 | Not recorded | 68fa6daa1e209131cea3b507 | | FL-53 | 0.201 | Not recorded | 68fa6daa1e209131cea3b507 | | MMCE | 0.203 | Not recorded | 68fa6daa1e209131cea3b507 | | L1 KDE-XE | 0.223 | Not recorded | 68fa6daa1e209131cea3b507 | ### L1 KDE-XE L1 ECEKDE: 0.077. Result ID: 698fbbb53a6c314849fb8065. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Wide-ResNet-28-10","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### FL-53 L1 ECEKDE: 0.081. Result ID: 698fbbb53a6c314849fb806b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"DenseNet-40","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### FL-53 L1 ECEKDE: 0.081. Result ID: 698fbbb53a6c314849fb806c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"DenseNet-40","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### XE L1 ECEKDE: 0.083. Result ID: 698fbbb53a6c314849fb805f. [Reporting paper]() Recorded evaluation configuration: {"Model":"Wide-ResNet-28-10","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### L1 KDE-XE L1 ECEKDE: 0.09. Result ID: 698fbbb53a6c314849fb806d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"DenseNet-40","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### L1 KDE-XE L1 ECEKDE: 0.091. Result ID: 698fbbb53a6c314849fb8066. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Wide-ResNet-28-10","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### XE L1 ECEKDE: 0.098. Result ID: 698fbbb53a6c314849fb8060. [Reporting paper]() Recorded evaluation configuration: {"Model":"Wide-ResNet-28-10","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### XE L1 ECEKDE: 0.104. Result ID: 698fbbb53a6c314849fb8067. [Reporting paper]() Recorded evaluation configuration: {"Model":"DenseNet-40","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### L1 KDE-XE L1 ECEKDE: 0.124. Result ID: 698fbbb53a6c314849fb806e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"DenseNet-40","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### L1 KDE-XE L1 ECEKDE: 0.126. Result ID: 698fbbb53a6c314849fb8055. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"ResNet-110","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### XE L1 ECEKDE: 0.131. Result ID: 698fbbb53a6c314849fb8068. [Reporting paper]() Recorded evaluation configuration: {"Model":"DenseNet-40","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### XE L1 ECEKDE: 0.132. Result ID: 698fbbb53a6c314849fb8057. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-110 (SD)","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### XE L1 ECEKDE: 0.133. Result ID: 698fbbb53a6c314849fb804f. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-110","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### MMCE L1 ECEKDE: 0.133. Result ID: 698fbbb53a6c314849fb8069. [Reporting paper]() Recorded evaluation configuration: {"Model":"DenseNet-40","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### FL-53 L1 ECEKDE: 0.138. Result ID: 698fbbb53a6c314849fb8053. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-110","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### MMCE L1 ECEKDE: 0.143. Result ID: 698fbbb53a6c314849fb8061. [Reporting paper]() Recorded evaluation configuration: {"Model":"Wide-ResNet-28-10","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### FL-53 L1 ECEKDE: 0.147. Result ID: 698fbbb53a6c314849fb8063. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Wide-ResNet-28-10","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### MMCE L1 ECEKDE: 0.155. Result ID: 698fbbb53a6c314849fb8062. [Reporting paper]() Recorded evaluation configuration: {"Model":"Wide-ResNet-28-10","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### MMCE L1 ECEKDE: 0.155. Result ID: 698fbbb53a6c314849fb806a. [Reporting paper]() Recorded evaluation configuration: {"Model":"DenseNet-40","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### FL-53 L1 ECEKDE: 0.156. Result ID: 698fbbb53a6c314849fb805b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-110 (SD)","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### L1 KDE-XE L1 ECEKDE: 0.163. Result ID: 698fbbb53a6c314849fb8056. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"ResNet-110","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### MMCE L1 ECEKDE: 0.164. Result ID: 698fbbb53a6c314849fb8059. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-110 (SD)","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### XE L1 ECEKDE: 0.17. Result ID: 698fbbb53a6c314849fb8050. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-110","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### MMCE L1 ECEKDE: 0.171. Result ID: 698fbbb53a6c314849fb8051. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-110","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### FL-53 L1 ECEKDE: 0.171. Result ID: 698fbbb53a6c314849fb8054. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-110","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### XE L1 ECEKDE: 0.172. Result ID: 698fbbb53a6c314849fb8058. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-110 (SD)","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### FL-53 L1 ECEKDE: 0.177. Result ID: 698fbbb53a6c314849fb8064. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Wide-ResNet-28-10","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### L1 KDE-XE L1 ECEKDE: 0.178. Result ID: 698fbbb53a6c314849fb805d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"ResNet-110 (SD)","Temperature Scaling":"Pre T"}. This record does not assert independent reproduction. ### MMCE L1 ECEKDE: 0.196. Result ID: 698fbbb53a6c314849fb8052. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-110","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### FL-53 L1 ECEKDE: 0.201. Result ID: 698fbbb53a6c314849fb805c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-110 (SD)","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### MMCE L1 ECEKDE: 0.203. Result ID: 698fbbb53a6c314849fb805a. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-110 (SD)","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. ### L1 KDE-XE L1 ECEKDE: 0.223. Result ID: 698fbbb53a6c314849fb805e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"ResNet-110 (SD)","Temperature Scaling":"Post T"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.