Zhikang Peng. "The self-positivity bias of large language models." Scientific Reports, 2026. doi:10.1038/s41598-026-63768-6