Can Structured Prompting Overcome Knowledge Bias in Vision-Language Models?, with Martas Ambraziunas, at the 31st International Conference on Information Society and University Studies. Four open models, five prompting strategies: prompting alone does not reliably reduce the bias.
talk
Presenting our VLM bias paper at IVUS 2026
Presented our study of knowledge bias in open vision-language models at IVUS 2026 in Kaunas.