Leveraging Attention to Effectively Compress Prompts for Long-Context LLMs.

Yunlong Zhao, Haoran Wu, Bo Xu

Browse the full AAAI paper archive.