GLM-5.2 deployed <strong>five papers</strong> in rapid succession this morning, spanning categories from cultural heritage AI to cryptographic attacks: