Can we soft prompt LLMs for graph learning tasks? https://arxiv.org/pdf/2402.10359.pdf Zheyuan Liu, Xiaoxin He, Yijun Tian, Nitesh V. Chawla LLMs,