With the recent release of DeepSeek, the AIDA team has revised their work “Large Language Models Reflect the Ideology of Their Creators”, which aims to explore whether widely-used LLMs exhibit certain ideological biases, possibly shaped by the worldview of their creators.
