Knowledge is power.
I have mainly works in the AI field.
Currently I have +40 citations. Ainât much, but honest work. Google Scholar
Research Briefs
Creating a Large Clean Web Corpus for Turkish (2025) Introduces a massive, high-quality dataset curated from web crawls specifically for Turkish NLP. The work focuses on rigorous cleaning and filtering pipelines to provide a robust foundation for training the next generation of Turkish LLMs.
Introducing CosmosGPT: Monolingual Training for Turkish Language Models (2024) Explores the efficacy of training Large Language Models from scratch on purely Turkish data. This research demonstrates that specialized monolingual training can outperform general multilingual models on local linguistic nuances and cultural context.
Optimizing Large Language Models for Turkish: New Methodologies in Corpus Selection and Training (2024) A deep dive into the technical optimizations required for the Turkish language, covering specific tokenization strategies and the impact of varied corpus selection on model performance and reasoning capabilities.
Performance Comparison of Turkish Language Models (2024) A comprehensive benchmarking study evaluating various Turkish-centric models. This paper provides a standardized framework for measuring success in tasks like sentiment analysis, summarization, and question-answering within the Turkish language ecosystem.
Cosmos-LLaVA: Chatting with the Visual (2024) Expands the Cosmos framework into the multimodal domain. This work integrates visual understanding with Turkish language processing, allowing for sophisticated âimage-to-textâ conversations and visual reasoning in a native Turkish context.