Papers by Krzysztof Jurkiewicz
Challenging America: Modeling language in longer time scales (2022.findings-naacl)
Copied to clipboard
Jakub Pokrywka, Filip Graliński, Krzysztof Jassem, Karol Kaczmarek, Krzysztof Jurkiewicz, Piotr Wierzchon
| Challenge: | a dominant approach to solving NLP tasks is pre-training a large neural language model and fine-tuning the model for specific tasks. |
| Approach: | They propose a challenge to train and fine-tune large Transformer models for historical texts . they pre-trained a RoBERTa model from scratch from the historical texts and evaluate them on benchmarks . |
| Outcome: | The proposed ML task is based on OCR-ed clippings from the Chronicling America portal. |
Oddballness: universal anomaly detection with language models (2025.coling-main)
Copied to clipboard
| Challenge: | a new method to detect anomalies in texts uses a metric called oddballness . the method considers probabilities generated by a language model but not low-likelihood tokens . |
| Approach: | They propose a method to detect anomalies in texts using unsupervised language models . they define oddballness as a function that measures how strange a given token is . |
| Outcome: | The proposed method is better than state-of-the-art models for grammatical error detection tasks. |