# OOD Detection on WMT15 En-Fr nt2014 OOD ID (test) Source: [OOD Detection on WMT15 En-Fr nt2014 OOD ID (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:26:12.454Z Evaluation of Out-of-Distribution (OOD) detection performance using the WMT15 English-to-French translation task configuration, specifically comparing In-Distribution data against the nt2014 OOD set on the test split, reporting AUROC and FPR95. 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 | OOD Detection | | Dataset (including any stated split) | WMT15 En-Fr nt2014 OOD ID (test) | | Metric | AUROC | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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 - [AUROC](): Higher is better; unit: PERCENT. - [FPR@95](): Lower is better; unit: PERCENT. ## Recorded results | Method | AUROC | Result date | Report paper | | --- | --- | --- | --- | | AP-OOD | 89.91 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | AP-OOD | 89.06 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Binary logits | 88.39 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Deep SAD | 88.27 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Binary logits | 87.39 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Relative Mahalanobis | 84.76 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Relative Mahalanobis | 84.17 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Entropy | 71.01 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | AP-OOD | 64.55 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Perplexity | 59.21 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | KNN | 58.99 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | AP-OOD | 56.44 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | KNN | 53.27 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Deep SAD | 51.65 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Mahalanobis | 49.7 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Mahalanobis | 49.08 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Deep SVDD | 48.03 | Not recorded | 698575e3c97a0b3e3b5d5f24 | | Deep SVDD | 46.7 | Not recorded | 698575e3c97a0b3e3b5d5f24 | ### AP-OOD AUROC: 89.91. Result ID: 6998ec605698665ec450e7f0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"OOD level":"Input","Learning paradigm":"Supervised","Backbone":"Transformer (base)"}. This record does not assert independent reproduction. ### AP-OOD AUROC: 89.06. Result ID: 6998ec605698665ec450e7f4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"OOD level":"Output","Learning paradigm":"Supervised","Backbone":"Transformer (base)"}. This record does not assert independent reproduction. ### Binary logits AUROC: 88.39. Result ID: 6998ec605698665ec450e7ed. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"OOD level":"Input","Learning paradigm":"Supervised","Backbone":"Transformer (base)"}. This record does not assert independent reproduction. ### Deep SAD AUROC: 88.27. Result ID: 6998ec605698665ec450e7ef. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"OOD level":"Input","Learning paradigm":"Supervised","Backbone":"Transformer (base)"}. This record does not assert independent reproduction. ### Binary logits AUROC: 87.39. Result ID: 6998ec605698665ec450e7f1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"OOD level":"Output","Learning paradigm":"Supervised","Backbone":"Transformer (base)"}. This record does not assert independent reproduction. ### Relative Mahalanobis AUROC: 84.76. Result ID: 6998ec605698665ec450e7ee. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"OOD level":"Input","Learning paradigm":"Supervised","Backbone":"Transformer (base)"}. This record does not assert independent reproduction. ### Relative Mahalanobis AUROC: 84.17. Result ID: 6998ec605698665ec450e7f2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"OOD level":"Output","Learning paradigm":"Supervised","Backbone":"Transformer (base)"}. This record does not assert independent reproduction. ### Entropy AUROC: 71.01. Result ID: 6998ec605698665ec450e771. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Output","Unsupervised":true}. This record does not assert independent reproduction. ### AP-OOD AUROC: 64.55. Result ID: 6998ec605698665ec450e775. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Output","Unsupervised":true}. This record does not assert independent reproduction. ### Perplexity AUROC: 59.21. Result ID: 6998ec605698665ec450e770. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Output","Unsupervised":true}. This record does not assert independent reproduction. ### KNN AUROC: 58.99. Result ID: 6998ec605698665ec450e773. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Output","Unsupervised":true}. This record does not assert independent reproduction. ### AP-OOD AUROC: 56.44. Result ID: 6998ec605698665ec450e745. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Input","Unsupervised":true}. This record does not assert independent reproduction. ### KNN AUROC: 53.27. Result ID: 6998ec605698665ec450e743. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Input","Unsupervised":true}. This record does not assert independent reproduction. ### Deep SAD AUROC: 51.65. Result ID: 6998ec605698665ec450e7f3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"OOD level":"Output","Learning paradigm":"Supervised","Backbone":"Transformer (base)"}. This record does not assert independent reproduction. ### Mahalanobis AUROC: 49.7. Result ID: 6998ec605698665ec450e772. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Output","Unsupervised":true}. This record does not assert independent reproduction. ### Mahalanobis AUROC: 49.08. Result ID: 6998ec605698665ec450e742. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Input","Unsupervised":true}. This record does not assert independent reproduction. ### Deep SVDD AUROC: 48.03. Result ID: 6998ec605698665ec450e774. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Output","Unsupervised":true}. This record does not assert independent reproduction. ### Deep SVDD AUROC: 46.7. Result ID: 6998ec605698665ec450e744. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Transformer (base)","Training Data":"WMT15 En-Fr","Evaluation Side":"Input","Unsupervised":true}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.