Researchers have developed CorPipe 26, a system that won the CRAC 2026 Shared Task on Multilingual Coreference Resolution. This improved system introduces a variant capable of predicting empty nodes alongside mentions and coreference links within a single model. CorPipe 26 demonstrated superior performance, outperforming other submissions in both the LLM and unconstrained tracks by significant margins. AI
IMPACT Advances multilingual coreference resolution, potentially improving cross-lingual understanding in NLP applications.
RANK_REASON Academic paper detailing a new system and its performance on a specific task.
AI-generated summary · Google Gemini · from 3 sources. How we write summaries →
