Please use this identifier to cite or link to this item: https://repository.hneu.edu.ua/handle/123456789/39401
Title: Small Language Models for perplexity-based text classification
Authors: Gorokhovatskyi O.
Keywords: Small Language Models (SLMs)
Text Classification
Perplexity
AI-generated Content
Gemma 3
Llama 3.2
Convolutional Neural Networks (CNN)
Ukrainian Language
Issue Date: 2026
Citation: Gorokhovatskyi O. Small Language Models for perplexity-based text classification / O. Gorokhovatskyi // Сучасні інформаційні системи та технології в цифровому суспільстві : матеріали Міжнародної науково-практичної конференції тези доповідей, 16 – 17 квітня 2026 р. : тези допов. – Харків : ХНЕУ імені Семена Кузнеця, 2026. – С. 86.
Abstract: The research explores the use of Small Language Models (SLMs), specifically those with 1B parameters or fewer, to classify text as either AI-generated or human-written in the Ukrainian language. By leveraging perplexity – a measure of a model's uncertainty – as a key feature, the study evaluates the performance of models like Gemma 3 and Llama 3.2. The authors further propose a classification method using Convolutional Neural Networks (CNN) trained on token-level probability vectors, achieving accuracy rates up to 0.8559.
URI: https://repository.hneu.edu.ua/handle/123456789/39401
Appears in Collections:Статті (ІКТ)

Files in This Item:
File Description SizeFormat 
SMALL LANGUAGE MODELS FOR PERPLEXITYBASED TEXT CLASSIFICATION.pdf418,51 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.