A Conditional No-Hair Theorem for Transformers: How Language Models Forget Their Prompt By Shuyang Xiang · HF Daily Papers · Oct 16, 2026 HF Daily Papers' note score 5 Categories: Research