6 papers
Precision--Recall Trade-off on the Inference Level for LLMs in Grammatical Error Correction
Ryszard Staruch, Filip Graliński
Large language models (LLMs) have recently been adopted for Grammatical Error Correction (GEC), achieving strong performance via both full fine-tuning and prompt-based approaches.…
Adapting LLMs for Minimal-edit Grammatical Error Correction
Ryszard Staruch, Filip Graliński, Daniel Dzienisiewicz
Decoder-only large language models have shown superior performance in the fluency-edit English Grammatical Error Correction, but their adaptation for minimal-edit English GEC is st…
LLMzSzŁ: a comprehensive LLM benchmark for Polish
Krzysztof Jassem, Michał Ciesiółka, Filip Graliński +5
This article introduces the first comprehensive benchmark for the Polish language at this scale: LLMzSzŁ (LLMs Behind the School Desk). It is based on a coherent collection of Poli…
Oddballness: universal anomaly detection with language models
Filip Graliński, Ryszard Staruch, Krzysztof Jurkiewicz
We present a new method to detect anomalies in texts (in general: in sequences of any data), using language models, in a totally unsupervised manner. The method considers probabili…
POLygraph: Polish Fake News Dataset
Daniel Dzienisiewicz, Filip Graliński, Piotr Jabłoński +3
This paper presents the POLygraph dataset, a unique resource for fake news detection in Polish. The dataset, created by an interdisciplinary team, is composed of two parts: the "fa…
Two Approaches to Diachronic Normalization of Polish Texts
Kacper Dudzic, Filip Graliński, Krzysztof Jassem +2
This paper discusses two approaches to the diachronic normalization of Polish texts: a rule-based solution that relies on a set of handcrafted patterns, and a neural normalization…