# Coreference Resolution on English OntoNotes 5.0 (test) Source: [Coreference Resolution on English OntoNotes 5.0 (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:51:19.638Z Evaluation of Coreference Resolution performance on the test split of the English OntoNotes 5.0 dataset, reporting MUC, B3, and CEAF metrics. 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 | Coreference Resolution | | Dataset (including any stated split) | English OntoNotes 5.0 (test) | | Metric | MUC Precision | | Unit | PERCENT | | Direction | Higher 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 - [MUC Precision](): Higher is better; unit: PERCENT. - [MUC Recall](): Higher is better; unit: PERCENT. - [MUC F1 Score](): Higher is better; unit: PERCENT. - [B3 Precision](): Higher is better; unit: PERCENT. - [B3 Recall](): Higher is better; unit: PERCENT. - [B3 F1 Score](): Higher is better; unit: PERCENT. - [CEAF phi4 Precision](): Higher is better; unit: PERCENT. - [CEAF phi4 Recall](): Higher is better; unit: PERCENT. - [CEAF phi4 F1 Score](): Higher is better; unit: PERCENT. - [Average F1 Score](): Higher is better; unit: PERCENT. - [CoNLL F1](): Higher is better; unit: PERCENT. - [LEA Recall](): Higher is better; unit: PERCENT. - [LEA Precision](): Higher is better; unit: PERCENT. - [LEA F1 Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | MUC Precision | Result date | Report paper | | --- | --- | --- | --- | | corefqa | 88.6 | Not recorded | 68fa70e71e209131cea4dd41 | | LINGMESS | 88.1 | Not recorded | 68fa6e911e209131cea40b6a | | link-append | 87.4 | Not recorded | 68fa70e71e209131cea4dd41 | | s2e | 86.6 | Not recorded | 68fa6e911e209131cea40b6a | | s2e + Longformer-Large | 86.5 | Not recorded | 68fa6a0f1e209131cea25ded | | Kirstain et al. | 86.5 | Not recorded | 68fa6c401e209131cea33065 | | ASP | 86.1 | Not recorded | 68fa70e71e209131cea4dd41 | | c2f + Longformer-Large | 86 | Not recorded | 68fa6a0f1e209131cea25ded | | Xu and Choi | 85.9 | Not recorded | 68fa6c401e209131cea33065 | | Joshi et al. | 85.8 | Not recorded | 68fa6c401e209131cea33065 | | c2f + SpanBERT-Large | 85.7 | Not recorded | 68fa6a0f1e209131cea25ded | | JOSHI-REPLICA | 85.2 | Not recorded | 68fa6c401e209131cea33065 | | s2e | 85.2 | Not recorded | 68fa70e71e209131cea4dd41 | | LingMess | 85.1 | Not recorded | 68fa70e71e209131cea4dd41 | | CAW-coref | 85.1 | Not recorded | 68fa70e71e209131cea4dd41 | | c2f + Longformer-Base | 85 | Not recorded | 68fa6a0f1e209131cea25ded | | wl-coref | 84.9 | Not recorded | 68fa6c401e209131cea33065 | | WL+ | 84.8 | Not recorded | 68fa70e71e209131cea4dd41 | ### corefqa MUC Precision: 88.6. Result ID: 69915fca3a6c31484907b68b. [Reporting paper]() Recorded evaluation configuration: {"LM calls":"O(n²)","params.":"340M","Link compl.":"/"}. This record does not assert independent reproduction. ### LINGMESS MUC Precision: 88.1. Result ID: 699036683a6c314849fef134. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### link-append MUC Precision: 87.4. Result ID: 69915fca3a6c31484907b68a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"LM calls":"O(n)","params.":"13B","Link compl.":"/"}. This record does not assert independent reproduction. ### s2e MUC Precision: 86.6. Result ID: 699036683a6c314849fef133. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### s2e + Longformer-Large MUC Precision: 86.5. Result ID: 698e0914ffc3901de09a49aa. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Architecture":"s2e","Backbone":"Longformer-Large"}. This record does not assert independent reproduction. ### Kirstain et al. MUC Precision: 86.5. Result ID: 698efb663a6c314849f4415f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"year":"2021"}. This record does not assert independent reproduction. ### ASP MUC Precision: 86.1. Result ID: 69915fca3a6c31484907b68c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"LM calls":"O(n)","params.":"11B","Link compl.":"/"}. This record does not assert independent reproduction. ### c2f + Longformer-Large MUC Precision: 86. Result ID: 698e0914ffc3901de09a49a9. [Reporting paper]() Recorded evaluation configuration: {"Architecture":"c2f","Backbone":"Longformer-Large"}. This record does not assert independent reproduction. ### Xu and Choi MUC Precision: 85.9. Result ID: 698efb663a6c314849f4415e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"year":"2020"}. This record does not assert independent reproduction. ### Joshi et al. MUC Precision: 85.8. Result ID: 698efb663a6c314849f4415c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"year":"2020"}. This record does not assert independent reproduction. ### c2f + SpanBERT-Large MUC Precision: 85.7. Result ID: 698e0914ffc3901de09a49a7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Architecture":"c2f","Backbone":"SpanBERT-Large"}. This record does not assert independent reproduction. ### JOSHI-REPLICA MUC Precision: 85.2. Result ID: 698efb663a6c314849f4415d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### s2e MUC Precision: 85.2. Result ID: 69915fca3a6c31484907b68e. [Reporting paper]() Recorded evaluation configuration: {"LM calls":"1","params.":"355M","Link compl.":"O(n⁴)"}. This record does not assert independent reproduction. ### LingMess MUC Precision: 85.1. Result ID: 69915fca3a6c31484907b68d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"LM calls":"1","params.":"355M","Link compl.":"O(n⁴)"}. This record does not assert independent reproduction. ### CAW-coref MUC Precision: 85.1. Result ID: 69915fca3a6c31484907b68f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"LM calls":"1","params.":"355M","Link compl.":"O(n²)"}. This record does not assert independent reproduction. ### c2f + Longformer-Base MUC Precision: 85. Result ID: 698e0914ffc3901de09a49a8. [Reporting paper]() Recorded evaluation configuration: {"Architecture":"c2f","Backbone":"Longformer-Base"}. This record does not assert independent reproduction. ### wl-coref MUC Precision: 84.9. Result ID: 698efb663a6c314849f44160. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### WL+ MUC Precision: 84.8. Result ID: 69915fca3a6c31484907b690. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"LM calls":"1","params.":"355M","Link compl.":"O(n²)"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.