Papers by Paweł Morawiecki
Can Models Help Us Create Better Models? Evaluating LLMs as Data Scientists (2026.findings-eacl)
Copied to clipboard
| Challenge: | Current benchmarks assess LLMs on more isolated capabilities, such as language understanding and question-answering. |
| Approach: | They propose a benchmark to evaluate the ability of large language models (LLMs) to perform feature engineering. |
| Outcome: | The proposed benchmark evaluates the ability of large language models to perform feature engineering, a critical and knowledge-intensive task in data science. |
Deep Neural Networks for Coreference Resolution for Polish (L18-1)
Copied to clipboard
| Challenge: | Existing deep neural networks for coreference resolution for Polish have been used to resolve textual fragments that refer to the same entity in the discourse world. |
| Approach: | They propose a system combining the best deep neural architecture and sieve-based coreference resolvers ordered from most to least precise to achieve the highest results. |
| Outcome: | The proposed system improves the state of the art for Polish by 0.53 F1 points, reaching 81.23 points of the CoNLL metric. |