NileUn at AbjadGenEval Shared Task: Contrastive Learning with Stacking Ensemble for Efficient Arabic AI-Generated Text Detection

Résumé

We present a computationally efficient approach for detecting AI-generated Arabic text as part of the AbjadGenEval shared task.Our method combines Supervised Contrastive Learning with a Stacking Ensemble of AraBERT and XLM-RoBERTa models.Our training pipeline progresses through three stages: (1) standard fine-tuning without contrastive loss, (2) adding supervised contrastive loss for better embeddings, and (3) further fine-tuning on diverse generation styles.On our held-out test split, the stacking ensemble achieves F1=0.983 before fine-tuning.On the official workshop test data, our system achieved 4th place with F1=0.782, demonstrating strong generalization using only encoder-based transformers without requiring large language models.Our implementation is publicly available.1

Citer ce document

Accès au document

Texte intégral en lecture en ligne, réservé aux abonnés SPHAERO et aux membres de l'institution. Se connecter

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0