From taja.kuzman at ijs.si Thu Feb 12 14:20:51 2026
From: taja.kuzman at ijs.si (=?UTF-8?Q?Taja_Kuzman_Punger=C5=A1ek?=)
Date: Thu, 12 Feb 2026 14:20:51 +0100
Subject: [CLASSLA] CLASSLA LLM Evaluation Dashboard for South Slavic
Languages
Message-ID: <64c906fa-f441-4631-8231-0f2cf69da407@ijs.si>
CLASSLA Mailing List
Dear all,
Are you wondering which large language model performs the best on South
Slavic languages and dialects? Or what is the performance of models for
sentiment, topic, genre classification and commonsense reasoning tasks,
especially compared to previous state-of-the-art models that have been
especially trained on these tasks?
At CLASSLA and CLARIN.SI, we have now set up an interactive dashboard
that allows you to search through the results of our evaluation of large
language models on various tasks. The models are evaluated on carefully
manually-annotated datasets most of which we developed in various
previous projects. You are warmly invited to visit the dashboard here:
https://www.clarin.si/classla-llm-dashboard/
For more details:
- Read the paper on the benchmarking experiments and results "State of
the Art in Text Classification for South Slavic Languages: Fine-Tuning
or Prompting?" (Taja Kuzman Pungeršek, Peter Rupnik, Ivan Porupski, Vuk
Dinić, Nikola Ljubešić, 2025): https://arxiv.org/abs/2511.07989
- Consult the code for the experiments published on GitHub:
https://github.com/TajaKuzman/Benchmarking-Text-Classification-on-South-Slavic/
**
Best wishes,
Taja, Nikola, and many other CLASSLAers
CLASSLA: The Knowledge Centre for South Slavic Languages
CLARIN.SI
Jožef Stefan Institute
Jamova cesta 39, Ljubljana
Slovenia
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
-------------- next part --------------
A non-text attachment was scrubbed...
Name: EgnHzq0OLDrCZSz9.png
Type: image/png
Size: 174960 bytes
Desc: not available
URL: