Coreference Resolution on WikiCoref
73.2Average F1 ScoreImCoref-CeSgpt4
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ImCoref-CeSgpt4evaluation_type=out-of-domain, reasoning_augmented_component=GPT-4, train_dataset=OntoNotes2025.10 | 73.2 | 83.6 | 69.8 | — | |
| InstructGPTtype=pure LLM-based2025.10 | 72.9 | — | — | — | |
| ImCoref-CeSdpskevaluation_type=out-of-domain, reasoning_augmented_component=DeepSeek, train_dataset=OntoNotes2025.10 | 72.4 | 82.4 | 69.2 | — | |
| ImCoref-CeSqwenevaluation_type=out-of-domain, reasoning_augmented_component=Qwen, train_dataset=OntoNotes2025.10 | 71.3 | 81.8 | 67.3 | — | |
| GPT-3.5-turbotype=pure LLM-based2025.10 | 70.8 | — | — | — | |
| ImCorefevaluation_type=out-of-domain, train_dataset=OntoNotes2025.10 | 67.6 | 80.6 | 66.7 | — | |
| Maverickmesevaluation_type=out-of-domain, train_dataset=OntoNotes2025.10 | 66.8 | 81.5 | 65.4 | — | |
| Dual-cacheevaluation_type=out-of-domain, train_dataset=OntoNotes2025.10 | 63 | 72.1 | 62.1 | — | |
| LingMessevaluation_type=out-of-domain, train_dataset=OntoNotes2025.10 | 62.6 | — | — | — | |
| longdocevaluation_type=out-of-domain, train_dataset=OntoNotes2025.10 | 60.1 | — | — | — |