Towards personalized language models: significant advancements at the cost of emerging risks
M Gniewkowski, PT Walkowiak, M Klonowski, T Walkowiak, P Syga
ECML-PKDD 2027
2027
Paweł Walkowiak
Doktorant
Zainteresowania badawcze
- nlp
- adversarial examples
- mechanistic interpetability
Grupy badawcze
Wykształcenie
- 2019
inżynier informatyki
Politechnika Wrocławska
- 2023
magister informatyki
Politechnika Wrocławska
Publikacje9
Resilience of Named Entity Recognition models against adversarial attacks
PT Walkowiak
International Journal of Electronics and Telecommunications 71(3), 1–6
2025
Is heterogeneous model soup tasty? A multidimensional evaluation of diverse model soups in language model alignment
D Motyka, PT Walkowiak, JA Moska, A Janz, B Żuk, K Seweryn
ICCS 2025 154–162
2025
An empirical assessment of LLM-based approaches to malicious webpage detection
G Mak, M Gniewkowski, PT Walkowiak, A Janz
ICCS 2025 162–176
2025
Precise language deception: XAI driven targeted adversarial examples with restricted knowledge
M Gniewkowski, PT Walkowiak, M Klonowski, T Walkowiak
ICCS 2025 49–60 · Cytowane przez 1
2025
Robustness of named entity recognition models
PT Walkowiak
System Dependability - Theory and Applications 306–315
2024
Do Not Trust Me: explainability against text classification
M Gniewkowski, PT Walkowiak, P Syga, M Klonowski, T Walkowiak
ECAI 2023 875–882 · Cytowane przez 1
2023