Hanna Shcharbakova

Hanna Shcharbakova · publications

Publications

2025

When Scale Meets Diversity: Evaluating Language Models on Fine-Grained Multilingual Claim Verification

Hanna Shcharbakova, Tatiana Anikina, Natalia Skachkova, Josef van Genabith

FEVER @ ACL 2025 · oral AnthologyCode
Open for a bit more detail
Table of macro-F1 scores. XLM-R reaches 57.7 on the test subset, 47.6 out of domain and 43.2 zero-shot, ahead of mBERT and mT5, and far ahead of Qwen 2.5, Mistral Nemo and Llama 3.1, which score between 9.6 and 16.1.
Macro-F1 across evaluation subsets for the Claim+Evidence configuration: specialised encoders against instruction-tuned LLMs.

Cross-Lingual Fact Verification: Analyzing LLM Performance Patterns across Languages

Hanna Shcharbakova, Tatiana Anikina, Natalia Skachkova, Josef van Genabith

RANLP 2025 · oral AnthologyCode
Open for a bit more detail
Two radar charts comparing Mistral Nemo, Llama 3.1 and Qwen 2.5 across around twenty languages each. Fine-tuned Qwen 2.5 covers the largest area on both subsets, with performance varying sharply by language.
Per-language macro-F1 for three LLMs, fine-tuned and prompted, on the test and zero-shot subsets.

LCTeam at SemEval-2025 Task 3: Multilingual Detection of Hallucinations and Overgeneration Mistakes Using XLM-RoBERTa

Araya Kiros Hailemariam, Jose Maldonado Rodriguez, Ezgi Başar, Roman Kovalev, Hanna Shcharbakova

SemEval 2025 @ ACL 2025 Anthology

2023

Code4ML: a large-scale dataset of annotated Machine Learning code

Published as Anna Scherbakova

PeerJ Computer Science · 2023 Paper

2022

Semantic Code Classification for Automated Machine Learning

Published as Anna Scherbakova

New in ML Workshop, NeurIPS 2021 · arXiv 2022 arXiv

Accuracy, syntactic complexity and task type at play in examination writing

Published as Anna Scherbakova

Studies in Corpus Linguistics · John Benjamins, 2022 Chapter

2021

A Corpus-Based Case Analysis on Syntactic Complexity in Russian ESL Learners’ Writing

Published as Anna Scherbakova

Conference proceedings · 2021

Graduate Student Conference in Learner Corpus Research, Inland Norway University of Applied Sciences.