Peng, Z. (2026). The self-positivity bias of large language models. Scientific Reports. https://doi.org/10.1038/s41598-026-63768-6