PoVisLE benchmark
AI & ML interests
None defined yet.
Recent Activity
PoVisLE benchmark
RefusEU dataset from "Multilingual Refusal Alignment for Safer Large Language Models" by A. Krasnodębska, W. Kusa, A. Lipani (ACL Findings 2026).
Polish Vision-Language Model based on PLLuM and LLaVA frameworks
Models and datasets from "PL-Guard: Benchmarking Language Model Safety for Polish" Krasnodębska, A., Seweryn, K., Łukasik, S., & Kusa, W. (2025)