9 papers
RDP LoRA: Geometry-Driven Identification for Parameter-Efficient Adaptation in Large Language Models
Yusuf Ãelebi, YaÄız Asker, Ãzay Ezerceli +4
Fine-tuning Large Language Models (LLMs) remains structurally uncertain despite parameter-efficient methods such as Low-Rank Adaptation (LoRA), as the layer-specific roles of inter…
A Hybrid Protocol for Large-Scale Semantic Dataset Generation in Low-Resource Languages: The Turkish Semantic Relations Corpus
Ebubekir Tosun, Mehmet Emin Buldur, Ãzay Ezerceli +1
We present a hybrid methodology for generating large-scale semantic relationship datasets in low-resource languages, demonstrated through a comprehensive Turkish semantic relations…
Beyond Cosine Similarity: Taming Semantic Drift and Antonym Intrusion in a 15-Million Node Turkish Synonym Graph
Ebubekir Tosun, Mehmet Emin Buldur, Ãzay Ezerceli +1
Neural embeddings have a notorious blind spot: they can't reliably tell synonyms apart from antonyms. Consequently, increasing similarity thresholds often fails to prevent opposite…
PARROT: Persuasion and Agreement Robustness Rating of Output Truth -- A Sycophancy Robustness Benchmark for LLMs
Yusuf Ãelebi, Ãzay Ezerceli, Mahmoud El Hussieni
This study presents PARROT (Persuasion and Agreement Robustness Rating of Output Truth), a robustness focused framework designed to measure the degradation in accuracy that occurs…
TurkColBERT: A Benchmark of Dense and Late-Interaction Models for Turkish Information Retrieval
Ãzay Ezerceli, Mahmoud El Hussieni, Selva TaÅ +4
Neural information retrieval systems excel in high-resource languages but remain underexplored for morphologically rich, lower-resource languages such as Turkish. Dense bi-encoders…
TurkEmbed: Turkish Embedding Model on NLI & STS Tasks
Ãzay Ezerceli, Gizem GümüÅçekiçci, TuÄba Erkoç +1
This paper introduces TurkEmbed, a novel Turkish language embedding model designed to outperform existing models, particularly in Natural Language Inference (NLI) and Semantic Text…