arXiv Artificial Intelligence· Ang Li, Sean McLeish, Haozhe Chen, Nimit Kalra, Zaiqian Chen, Artem Gazizov, Venkata Anoop Suhas Kumar Morisetty, Bhavya Kailkhura, Harshitha Menon, Zhuang Liu, Brian R. Bartoldson, Tom Goldstein, Sanae Lotfi, Micah Goldblum, Pavel Izmailov·· 3 小时前AI 评分52
LCLMs：面向长上下文推理的端到端上下文压缩
End-to-End Context Compression at Scale
AI 导读
研究团队提出 Latent Context Language Models（LCLMs），用编码器-解码器把长 token 序列压缩为更短的潜在嵌入，以缓解 KV cache 随上下文增长带来的显存瓶颈。
来源：arXiv Artificial Intelligence · arxiv.org