Advancing Natural Language Processing at METU through cutting-edge research in Large Language Models, Generative AI, and beyond.
Our research study titled "Impact of Bilingual Corpus on Pretraining Large Language Models: An Analysis for Turkish" has been accepted by the Multilingual Representation Learning Workshop at EMNLP 2026 and will be published soon.
Our research study titled "Detecting Gender-Motivated Hate Speech Across Domains with Contrastive Error Steering" has been accepted by the AI4GOOD Workshop at NeurIPS 2026 and will be published soon.
Our research study titled "Cross-Lingual Transfer for Machine Translation in Turkic Languages" has been published as an arXiv preprint. See details at arxiv.org
Our research study titled "OCRTurk: A Comprehensive OCR Benchmark for Turkish" has been accepted by the SIGTURK Workshop at EACL 2026 and published in its proceedings. See details at aclanthology.org
Our research study titled "RAGTurk: Best Practices for Retrieval Augmented Generation in Turkish" has been accepted by the SIGTURK Workshop at EACL 2026 and published in its proceedings. See details at aclanthology.org
Our research study titled "BIRDTurk: Adaptation of the BIRD Text-to-SQL Dataset to Turkish" has been accepted by the SIGTURK Workshop at EACL 2026 and published in its proceedings. See details at aclanthology.org
Our research study titled "TurkBench: A Benchmark for Evaluating Turkish Large Language Models" has been accepted by the SIGTURK Workshop at EACL 2026 and published in its proceedings. See details at aclanthology.org
Our project, titled “Social Network Modeling Using Persona-Based Large Language Models,” has been selected for funding under the TÜBİTAK ARDEB 1001 Programme’s First Call of 2026.
Our group is awarded a total of USD 80,000.00 in Google Cloud Credits specifically to support our research projects at Middle East Technical University.
We publish TurkBench at huggingface.co/spaces/TurkBench
Our research study titled "Evaluating the Quality of Benchmark Datasets for Low-Resource Languages: A Case Study on Turkish" has been published in the Proceedings of the Fourth Workshop on Generation, Evaluation and Metrics at ACL 2025. See details at aclanthology.org
Multilingual Representation Learning Workshop accepted our paper titled "Adapting Open-Source Generative Large Language Models for Low-Resource Languages: A Case Study for Turkish"
Published in Natural Language Processing, Cambridge University Press. Click here to see the article.